IEEE
You are not logged in, please sign in to edit > Log in / create account  

First-Hand:Philips Telephone Exchanges and Denmark 1990 - 1997

From GHN

(Difference between revisions)
Jump to: navigation, search
Line 52: Line 52:
 
<br>Early in 1992 Erik Andreasen called for me to give me a top secret assignment. I were to be project leader for an evaluation of whether Siemens HICOM PABX should be included in CTBD’s program of large PABXes as an alternative to L. M. Ericsson’s MD110. An engineer had been employed to take care of the details. He was so secret that he had no office at CTBD but at Siemens. Tage Fox Maule was also assigned to the project.<br>The three of us took part in meetings and were on a course in Munich. I was also in Berlin on a one-day trip to discuss the connection of “customer links” (2 Mbit/s PCM connections) to HICOM.<br>An exchange came to Denmark and was installed in the Lyngby exchange building. The two others studied it in depth. Several others were called in to work on sales preparation and sales opportunities. The discussions about its price were particularly difficult. Siemens failed to understand that we wanted an exchange without their special facilities at a price at the MD110 level, such that it could compete. They rather found that if we only sold it with all their bells and whistles it should be possible to ask a much higher price. But – the market would then also be very little.<br>During the autumn it was decided that CTBD and the corresponding division of Jutland Telephones should be combined under the head of the Jutland division, Søren Jensen. The two divisions had invaded the area of the other part and started a mutual price war. This was devastating for the earnings of Tele Danmark. The final blow was when a large order for PABXes to the public administration in Copenhagen went to the Jutland division at rock bottom prices for their Meridian exchange.<br>The combination led of course to a re-evaluation of the PABX offerings. It led to the marketing by Tele Danmark Business Division (TDBD) of the two systems they already had on their respective programmes. Siemens was notified that the work on HICOM would stop.<br>The liberalisation was already a couple of years old and Siemens choose to market their system directly. Thus they have probably learned that prices in Denmark were very low compared with several other markets and that nobody wanted to pay much for a PABX with basic facilities. The technical development and the competition were hard for all suppliers. From a price to the end user of about 6000 kr. (USD 1000) per extension in 1980 (a price which made no supplier rich) through about half of this in the late 1980es, the prices decreased to 1500 kr. (USD 250) per extension in the middle of the 1990es at the same time as there was still more individual adaptation of the systems. It didn’t really fit but that is what the suppliers, i.a. TDBD had to live with!<br>
 
<br>Early in 1992 Erik Andreasen called for me to give me a top secret assignment. I were to be project leader for an evaluation of whether Siemens HICOM PABX should be included in CTBD’s program of large PABXes as an alternative to L. M. Ericsson’s MD110. An engineer had been employed to take care of the details. He was so secret that he had no office at CTBD but at Siemens. Tage Fox Maule was also assigned to the project.<br>The three of us took part in meetings and were on a course in Munich. I was also in Berlin on a one-day trip to discuss the connection of “customer links” (2 Mbit/s PCM connections) to HICOM.<br>An exchange came to Denmark and was installed in the Lyngby exchange building. The two others studied it in depth. Several others were called in to work on sales preparation and sales opportunities. The discussions about its price were particularly difficult. Siemens failed to understand that we wanted an exchange without their special facilities at a price at the MD110 level, such that it could compete. They rather found that if we only sold it with all their bells and whistles it should be possible to ask a much higher price. But – the market would then also be very little.<br>During the autumn it was decided that CTBD and the corresponding division of Jutland Telephones should be combined under the head of the Jutland division, Søren Jensen. The two divisions had invaded the area of the other part and started a mutual price war. This was devastating for the earnings of Tele Danmark. The final blow was when a large order for PABXes to the public administration in Copenhagen went to the Jutland division at rock bottom prices for their Meridian exchange.<br>The combination led of course to a re-evaluation of the PABX offerings. It led to the marketing by Tele Danmark Business Division (TDBD) of the two systems they already had on their respective programmes. Siemens was notified that the work on HICOM would stop.<br>The liberalisation was already a couple of years old and Siemens choose to market their system directly. Thus they have probably learned that prices in Denmark were very low compared with several other markets and that nobody wanted to pay much for a PABX with basic facilities. The technical development and the competition were hard for all suppliers. From a price to the end user of about 6000 kr. (USD 1000) per extension in 1980 (a price which made no supplier rich) through about half of this in the late 1980es, the prices decreased to 1500 kr. (USD 250) per extension in the middle of the 1990es at the same time as there was still more individual adaptation of the systems. It didn’t really fit but that is what the suppliers, i.a. TDBD had to live with!<br>
  
=== The N4H group ===
+
=== The N4H group ===
  
<br>In 1992 the N4H group was started with Knud Solholt from Jutland Telephones as chairman. He was also the Danish representative in the ETSI technical committee TE5. This committee was responsible for ETS 300001 or NET4, the rules for terminals on the analogue switched telephone network (or PSTN, Public Switched Telephone Network). NET4 from 1991 was a catalogue over the different rules in Europe, as national devia-tions were OK then. The differences were only removed when a change was made to TBRs (Technical Basis for Regulation), and TE5 had started a project to harmonise these rules. I became part of the N4H group because I had been part of the ETSI project team PT26, NET4 editing group, in the autumn of 1990.<br>The N4H group should evaluate the papers from TE5 from a Danish viewpoint. The name meant of course NET4 Harmonisation. The papers started with a discussion of which requirements to the terminals it was possible to harmonise and then went on to what the harmonised values should be.<br>We took most of the European discussion very quietly. We let the big network operators carry the burden and only occasionally our discussions made Solholt try his influence in TE5.<br>One of the essential questions was whether the loop resistance could be harmonised? The loop resistance is the DC resistance a terminal in its off-hook state presents to the network, such that the network perceives the terminal as busy. The Danish requirement in NET4 was that below 16 mA loop current the voltage between the terminal connec-tion points should be below 6,4 V, and above 16 mA the DC resistance between the connection points had to be less than 400 Ohm. This value or a lower one could very well be harmonised in our opinion. Although it might be a tightening related to what was required in Denmark if it was reduced to e.g. 300 Ohm, there would be no problems getting terminals complying with the requirement if it was a harmonised value.<br>However, an item quite impossible to harmonise was the French requirement that a ter-minal had to limit the loop current to a maximum of 60 mA. They stated it was neces-sary as the current supply from the line circuits in the exchange would saturate them and cause a lower quality of the transmission if the current was higher. Everybody else found that this was a problem handled by the exchanges themselves, making the termi-nals responsible could only be considered a technical trade barrier.<br>The discussion in ETSI continued for a long time after our group had changed into the N4G group with the task to revise the Danish part of NET4. Only in November 1994 the ETR 075 was issued with the result of the efforts by ETSI to harmonise NET4.<br>The DECT group<br>Another group in which I participated from 1992 was the DECT group. DECT is Digital European Cordless Telephony and is a technique to allow many users to have each his cordless telephone network in use within a small area. DECT is not an independent net-work like the mobile telephone network, each DECT system works under a public or private (PABX) exchange or network of exchanges in the PSTN. It could be used e.g. by the owner of a PABX to give its extensions local mobility unhampered by a neighbour giving the same opportunity to his employees.<br>LME (L. M. Ericsson) had done early work in this area and had developed and had in test operation several systems according to the same principles, but at frequencies around 900 MHz. The standards were now so far that a range of frequencies around 1900 MHz had been reserved and various protocol questions had been settled. LME could supply systems according to the latest version of the standards and the Danish telecom administrations would like to test them in different applications.<br>An arrangement was made for equipment for four tests, one at each administration. In Jutland it was set up under a public exchange in Aalborg as an alternative to the copper pairs in the subscriber cables, In Southern Jutland it was set up under PABXes in the frontier town Padborg to give local mobility to the personnel of the transportation firms. On Funen it was set up under the PABX in the headquarters of the administration to give mobility to the employees. In the Copenhagen area it was set up in a shopping cen-tre in Hillerød under the public exchange. The handsets should work in parallel with people’s PSTN telephones, such that the users could have their telephones with them when they were shopping.<br>The DECT group prepared the tests and should evaluate the result. It was indeed started but died out due to the changes in the organisation (i.a. the fusion of the business divi-sions). It is my impression that there were never many customers to the application in Hillerød. Maybe because most people do not want to be called on the telephone when they are shopping – and the others use a proper mobile telephone, the use of which was not limited to one shopping centre!<br>Øberg leaves CTBD<br>Already before the fusion of the business divisions the “war” between them had caused heavy discussions within each administration. This also led to Øberg taking his leave. Some years later when I met him in Horsens he said it was a disagreement about what CTBD should do between him and the chairman of the board, the CEO of CT Jørgen Lindegaard, which had caused him to quit his job.<br>He was not unemployed for long. He told me that he was called by the CEO of Alcatel Kirk with an offer to replace him. It suited him better to become CEO of S. C. Sørensen in Randers (I think it was a family business). Thus Øberg was employed here and trav-elled between their sites in Ballerup and Horsens until the Alcatel operation in Horsens was closed down. Their PABX production could not any more compete with the large players on the market and Alcatel could in stead import PABXes to the Danish market.<br>Other parts of the production of Alcatel could also not continue, but they were bought by the employees and continued under the old name Kirk. It was the production and further development of transducers, i.e. microphone and telephone capsules for fixed and mobile telephones, a large export item as the employees were and are experts in this area. And it was the production and further development of telephone sets sold both in Denmark and abroad. Alcatel concentrated on the production of space electronics in Ballerup and the sales of telecom equipment from abroad.<br>I also left the fused TDBD, but not on my own initiative and far less dramatically. Small PABXes up to 40 extensions should now be sold by the main companies, so many em-ployees from TDBD were transferred to CT and Jutland Telephones. I was one of these although I never had worked with such small PABXes while in CTBD. I came to the CT Telecom Division in the spring of 1993.<br>ETSI STC BTC2<br>By January 1st 1992 those of the employees of CTBD who had from old times been em-ployed in line with public servants, should decide whether they would be employed on a contract in CTBD or return to “big CT”. I.a. Niels Kristiansen, who was in the PABX group, chose to return. He then had to leave the PABX group and I proposed – and got approved – that he was replaced by another CTBD employee, Torben Paaske.<br>Niels also had to leave the work in ETSI STC BTC2 (or BT2 as it was called then) where he participated along with Jørgen Hess Kristensen from JTBD, Jutland Telephone Business Division, (who was the Danish spokesman). I suggested that I should replace Niels and this was accepted. I participated in meetings in Munich May 1992, in Bath October 1992, Sophia Antipolis April 1993 and Oslo September 1993.<br>Munich is the home town of Siemens and CTBD had started discussions about the in-troduction of HICOM in Denmark with them a few months earlier. This should be kept a secret for our competitors in JTBD, so I kept mum towards Jørgen. Well, it corre-sponded very much to what I was used to from Philips. We were always very friendly towards people from our competition but kept mum about company secrets. And so did the others!<br>When I received the flight ticket, I also got a telecard (to use in pay phones) with a value of DEM (Deutsche Mark) 5 from Lufthansa. It had a Danish text urging to use the card during the trip. I did so, but during the walk to the meeting place I passed a shop on a corner of the main station advertising the buying and selling of telecards. This I would of course test and every price from DEM 1 upwards would be a profit! Thus, when the card had been used I visited the shop, the card was set in a reader (to check the remain-ing charge and to ensure it was a real telecard) while I muttered that a German telecard with Danish text had to be a rarity. Well, they offered 2 DEM for it which I gladly re-ceived.<br>The German Telecom, who had arranged the meeting, invited all participants on an eve-ning trip to Garmisch-Partenkirchen, where we had a good meal and enjoyed local per-formers – and a performance of some of us, who showed their muscle power by compet-ing about who could hold a filled beer glass in an outstretched hand for the longest time.<br>The meeting in Bath was called by British Telecom. I used the opportunity to postpone my trip home to Saturday, such that I could be a tourist Friday afternoon. Going by bus around in the town, visiting the Roman Baths and the Cathedral, the latter with i.a. the epitaph from one great actor to the other, de Garrick writing over James Quin with the potent final words: “Whate,er thy strength of body, force or thought; In nature’s happi-est mold however cast; To this complexion thou must come at last”.<br>Several years later I found these names in the book by Knud H. Thomsen “Røverne I Skotland” (The Scottish Robbers). But back in 1992 I had borrowed the book “Sarum” at the public library. This book is about Salisbury from the Ice Age to the present. This town and Stonehenge are not far from Bath and Bath is mentioned at about the time when the Roman Baths were built.<br>At the meeting in Oslo I was accompanied by my wife, Jytte. We made it a small holi-day, first a couple of days in the Swedish archipelago north of Gothenburg and after the meeting a Friday afternoon as tourists (which my wife had also been every day during the meetings). The meeting was well attended from Denmark, with 4 participants, i.a. Bent Qvist from Copenhagen Telephones just like me. This was because the last items in the standard for analogue leased lines should be decided at the meeting.<br>Friday Jytte and I made two trips to the woods North of Oslo. First we went by rail to a place with a beautiful lake which we circled, then by another railway to the Frogner hills. We made again a long walk and enjoyed then hot chocolate in a restaurant with a view to Oslo where we saw the boat to Copenhagen sail out.<br>Spring 1993, the Rouban Group<br>During the meeting of BTC2 in April 1993 I became chairman for a small sub-sub-group.<br>BTC2 was busy with standards for leased lines, the task I had helped planning in the autumn of 1991. The job was done in project team PT22V and it was then discussed in working party WPF of BTC2. My place was in WPA on transmission, but I got mixed up in the plenary discussions of the results of WPF. In any way the requirements to a terminal on an analogue leased line as to the power it might transmit outside the speech band was thoroughly discussed at the meeting in Sophia Antipolis.<br>The problem was that while the requirements to the connection, which the telecom ad-ministrations should comply with, were comprehensive in order to ensure that the cus-tomers got a connection of a certain quality (as regards attenuation, distortion, delay, noise etc.), the requirements to the terminals were governed by the Terminal Directive. It lists 7 essential requirements to which a terminal must comply, viz. 1) user safety, 2) safety for the telecom personnel, 3) EMC (no disturbances to and from other apparatus), 4) no harm to the network, 5) a reasonable use of radio frequencies, 6) cooperation with the network to make, keep and release connections and 7) cooperation with other termi-nals via the network in justified cases. Requirements 1), 2) and 3) shall be complied with through other standards and 5), 6) and 7) were irrelevant for terminals on leased lines (no radio transmission, the connection exists all the time and can not be changed from the terminal and 7) was to be considered for speech transmission only). This left only 4), no harm to the network, and what did that mean?<br>There were fanatics (“liberals”) who wanted to limit this to mean only physical damage, such as when a high voltage from a terminal could make network equipment fail. We obtained however agreement that it should also mean that a terminal was not allowed to send signals disturbing other users of the telecom networks. Such was the stance in April 1993. I had made a proposal ensuring that the signals sent could not via crosstalk (i.e. coupling to other wire pairs in the cable) disturb others. This was however not enough. During the meeting it was emphasised especially by the participants from Brit-ish Telecom that transmission systems where connections are multiplexed on a common path did not isolate the channels completely from each other. On a carrier system the channels are placed besides each other within a range of frequencies. If a channel sends too much power at the frequencies of a neighbouring channel it will disturb. The filters for each channel are not ideal, partly as a question of price, partly as a question of the technique. If a filter cuts off unnecessary frequencies too sharply it will distort the wanted frequencies close to the edge. All this was a quite justified viewpoint.<br>The discussion made the Chairman of BTC2, John Horrocks from British Telecom, pro-posed a group formed to present a proposal for requirements at the WPF meeting in July and he asked me at the meeting if I would chair this group.<br>I made a long pause. Yes, I would like to, but I had recently come to CTTD, Copenha-gen Telephones Telecom Division, (when CTBD and JTBD were joined into TDBD and small PABXes were moved to CTTD), and I had sensed the attitude to the work in ETSI. Thus it was clear for me that the job as chairman had to be handled without trav-elling. It was also only for a short time. The thoughts ran through my head until I saw a way and accepted the job. I collected names and fax numbers of the interested BTC2 members.<br>As soon as I was back home I described my ideas in a fax to the interested people. At the same time I coined the name The Rouban Group for “Requirements to OUt-BAnd Noise”. The idea was to send my proposal as far as it went by fax on May 1st, June 1st and July 1st, and then the interested parties could fax their comments by the 15th of each month. The schedule was to call for proposals in May, to make a draft by June and a final draft in July. There would just be time to get a yes or no to this before the WPF meeting in the end of July.<br>It was carried out as proposed. There was a good proposal from Holland and it was re-written to the standard formulation of ETSI. There were some comments but nothing disastrous in June, so it was largely this proposal which went out in July. It was unani-mously accepted (at least nobody had comments to the changes from June to July). Brit-ish Telecom even wrote that they were very satisfied and could support the proposal 100%. It also passed through the WPF meeting without amendments except for a few minor adjustments. One was that L. M. Ericsson rather late had observed that the im-pedance over which the power was measured was not the correct one outside the speech band. I had confirmed this in a comment to the final proposal and the change was intro-duced.<br>The problem was that inside the speech band a fixed impedance is used equal to the im-pedance at 1020 Hz, although the impedance changes over the band. When one moves outside the speech band and as here even up to 1 MHz the impedance deviates however too much from the impedance at 1020 Hz to keep it at this value.<br>The results from the Rouban-group were included in the requirements to terminals for analogue leased lines according to the ONP-LL directive, TBR15 and TBR17.<br>It is my impression that much standardisation work could be done as in the Rouban-group by contact via fax or e-mail, when people have first met and got to know each other. It is also my impression that the few meetings are well prepared by these means so everybody is fully oriented about all viewpoints before them.<br>The N4G-group<br>During 1993 the work on NET4 continued in the N4G-group. The job was to revise the Danish part of NET4 (or ETS 300001) to be part of a new version of the standard.<br>The first version had an exception for PABXes and there was in stead a reference to national regulations for them. These words would be removed, but NET4 had on several points values different from those in the Danish requirements to PABXes.<br>The PABX-group was by the way busy revising part 3 of these technical requirements, which treated their connection to analogue subscriber lines in the switched telephone network. I was editor on this task. We were ready a couple of months before the N4G-group was ready, but the Telecommunication Supervisory Board decided that the new version should not be issued while waiting for NET4. All these efforts for nothing, but the work was of course valuable for the work in the N4G-group.<br>One of the great differences of opinion within the group related to the leakage resistance of the terminal to earth. Circular 2 (requirements of the network to subscriber terminals in general on the switched network) said that the subscriber installation as a whole was allowed to have a minimum of 10 MΩ from each wire to earth. The first issue of NET4 required a minimum of 100 MΩ from each wire to earth for each terminal (several might be connected in parallel). At an early moment we had seen in the PABX-group that this was a too small value for a PABX when it should measure the polarity over the wires (used to check if the subscriber line was busy in the public exchange or not, in order to release the line as soon as an external caller had gone on hook). The PABX would require at least 10 MΩ, and as there might be more terminals in parallel con-nected to the NTP (Network Terminating Point) Circular 2 had to allow a correspond-ingly smaller leakage resistance of the total subscriber installation. Considering that the requirement to the subscriber cable was a minimum of 100kΩ from each wire to earth or 100 times less than required in NET4 of the subscriber terminal there was an ample margin to exploit. We had already in 1989 discussed this point with the technical com-mittee of the telephone administrations and received an accept of a minimum value in Circular 2 for the whole subscriber installation of 3 MΩ from each wire to earth. We PABX specialists used this value repeatedly in the discussions in the N4G-group, as this was our basis for requiring the value in NET4 changed from 100 to 10 MΩ, such that NET4 could also apply to PABXes.<br>We did not succeed completely. The revised text measures leakage resistance between short-circuited line terminal points, i.e. in the worst case half as much as from each point to earth, and this value was set to 10 MΩ in general. A considerable reduction compared with the first version where the same value would have been 50 MΩ. How-ever, the text was made acceptable due to an exception for PABXes. They were allowed a leakage resistance between the short-circuited terminal points and earth of minimum 5 MΩ.<br>This implied of course that the term PABX had to be defined in NET4. NET4 already had requirements to terminals connected in series, defined as terminals with one port towards the network and one towards a NET4 terminal, where the second terminal only receives power for its operation from the first terminal. We worked much on the defini-tion of a PABX and got as the result that a PABX was any other terminal with these two types of ports. Thus, if there were only two ports it was a requirement that the terminal on the second port was at least partly fed locally and if there were more than two ports it was by definition always a PABX.<br>The work was ready in the end of 1993 so it could be notified within the EC and could be sent to ETSI. It was, however, not included in the April 1994 version of NET4. It was only included in 1995 and the Telecommunication Supervisory Board issued a Dan-ish version of NET4, which only included the common text with the Danish values and the special Danish amendments such as the definition of a PABX.<br>The PANT-group<br>In 1994 I became part of a group preparing a buyer’s specification for an NT, Network Terminal, for ISDN PRA (Integrated Services Digital Network, Primary Rate Access), i.e. the unit installed by Tele Danmark (TDC) at customer premises to provide the inter-face to a 30 channel connection on a 2 Mbit/s subscriber line. The specification should of course be based on the standards from ETSI, and they covered the digital line, errors on it as well as the T interface, which is the interface to the customer’s equipment. The group was established by IU-A, the Infrastructure Committee, Access, of TDC.<br>But as always there were options, possibilities for choice, in the standards and our task was to choose which of these should apply to the NTs bought by TDC.<br>Chairman of the group was Ole Philbert from Jutland. Members were Erik Bisgaard and Poul Erik Pedersen from Jutland and Otto Kimer and me from Copenhagen.<br>The purpose of such a specification is of course to make common rules for the behav-iour of the equipment, such that we were not tied to use only NTs from LME under LME exchanges, or NTs from Siemens under the exchanges they had supplied. It should also be possible to buy NTs from a third supplier and use them under both types of ex-changes. The more the different network elements can be independently bought, the better prices can be obtained by TDC!<br>We had many meetings discussing the possible choices and their consequences. With my background I insisted from the start that the NT we should specify ought to be us-able also on “subscriber links”, 2 Mbit/s digital lines with switched telephony used as subscriber lines to digital PABXes, and on 2 Mbit/s structured leased lines, which we should provide as a part of our obligations according to the EC directive about leased lines (ONP-LL) and where I had participated in the planning of the standardisation work in ETSI PT9V during the autumn 1991. It turned out that the various requirements fitted together, such that this could be realised.<br>It also turned out that I got the job of collecting the contributions and reformulate them to the required application specifications to the ETSI standards. There were three stan-dards, one for general conditions (ambience etc.), one for the T-interface and one for the digital line.<br>The job was finished in January 1994 and the group was disbanded a couple of months later. In July 1994 TDC then issued their technical specification TDK-TS 900213 with our specifications.<br>I got mixed up in the use of the specifications later. Poul Erik sent me compliance statements, i.e. confirmations that the requirements were complied with, from LME and Siemens for evaluation. I remember that Siemens did not comply with one of the re-quirements, that the NT towards the customer should set the bits S(a5) and S(a6) used on the line to the value 1. Siemens had chosen to set them at 0. Back to the ETSI speci-fication for the T-interface: it required that the terminal should ignore these bit. We had made the requirement mandatory, but could accept the choice made by Siemens. If we later had to revise the specification we would probably drop our requirement.<br>Later while in the network division (after the old administrations had been replaced by functional divisions in TDC) my boss, Martin Gregersen, came with a data sheet for such an NT from Fujitsu. I said to him that he should ask Fujitsu to make a compliance statement according to TDK-TS 900213 as a basis for further discussions. I do not know whether this led to our buying these NTs.<br>Finally, in 1997, Poul Erik asked me if I still had the files from the past. They had been included in INFR, ISDN Functional Requirements, and these should now become TDK-TSes. Thus it would be good to have the original files. Luckily I had always made back-ups on floppies and had brought them with me from CTBD via CTTD to the network division. Thus I could go back in spite of the change of PC and operating system and send them to him. Another piece of luck was of course that the floppies could still be read although I had changed PC!<br>The PSTN-group<br>In 1994 I became part of a group specifying the monopoly interface in the switched ana-logue telephone network. This was probably another consequence of my work in ETSI during the autumn of 1990.<br>Our chairman was Erik Bisgaard from Jutland. The group was established by IU-A, the Infrastructure Committee – Access network.<br>The monopoly interface or - as it was later called - the NTP, Network Terminating Point, is the point where the TDC cable meets the subscriber installation. The liberalisa-tion meant that not only were the terminals, but also the whole installation at the sub-scriber premises, outside the TDC monopoly in making telephony connections. A de-scription was required of what TDC delivered in the NTP and of what we delivered be-tween two NTPs, corresponding to the network interface and the connection properties in the specifications for leased lines.<br>We had a good starting point in the old Circular 2 (from the 1970es), which had a sec-tion dealing with the subscriber installation. There was i.a. a requirement for a minimum loop current of 15,7 mA when the NTP was terminated in 600 Ω, and the distribution of the 600 Ω was described (400 Ω in the final terminal, 200 Ω in the subscriber installa-tion including serially connected terminals). This had to be modified as the requirements to terminals were changed since the 1970es (the old Circular 27), e.g. the requirement to the final terminal at loop currents below 16 mA was not any longer a maximum of 400 Ω, but a maximum of 6,4 V over its terminating points, and this had to be reflected in the properties of the NTP. I emphasised that as far as possible we should specify the requirements to be in line with Circular 2.<br>At the same time IU-T, the Infrastructure Committee – Transport Network, was busy on a similar job. They rather looked at how little TDC could provide in the NTP and would base their efforts at the work in ETSI to make a common European set of requirements to telecommunication terminals in TBR21, i.e. requirements ensuring that a terminal might be used throughout Europe. Their attitude was that the network should not prom-ise better properties than just required to serve a TBR21 terminal.<br>My standpoint – to which my group adhered – was that it was not the network which had to adapt to TBR21, but the TBR21 terminal which had to adapt to the network. Cir-cular 2 was the Danish starting point. It said e.g. 600 Ω terminal resistance in the NTP. If another network operator required 500 Ω terminal resistance and this became part of TBR21 it would be OK for us, but should not change our specifications for the network. If on the other hand it was sufficient for all other European network operators to require a loop resistance of 800 Ω, we should require of TBR21 that its requirement was our 600 Ω. Thus, our group insisted on maintaining the 600 Ω loop resistance over the NTP, while the other group would adhere to the requirement of maximum 500 Ω loop resis-tance in the terminal according to TBR21, and then we should not promise that it would be perceived as a loop (off hook) if there were more than 500 Ω loop resistance over the NTP.<br>This ended up, as one can guess, with two specifications with slightly different contents. The shape of our draft became the final one, but it was decided in IU that the figures from IU-T should be used, which they were. I got angry during this, when it was re-ported that the chairman of IU-T, Ole Jahn Svinding from Jutland, had said in an IU meeting that there was agreement about the figures from IU-T. This was blatantly incor-rect. But in the final discussions Erik Bisgaard was under a large pressure, and therefore the figures from IU-T were included in the final specification, issued as TDK-TS 900221.<br>Of course I could not help laughing later when the figures in TBR21 were modified (and TBR21 was by the way not finally approved as late as in the spring of 1997), meaning that there is no more a reference for the figures in TDK-TS 900221!<br>ETSI STC TE5<br>A consequence of the work in the N4G-group and the PSTN-group was that I also be-came part of ETSI STC TE5. However, I took only part in one meeting, in Mainz in October 1994, invited by the German Telecom.<br>At this time TE5 was busy finishing TBR21, a common European NET4. It applied to non-voice equipment, a corresponding TBR37 should be issued for telephone sets with TBR38 valid for the speech transmission itself.<br>There had already been lengthy discussions of which requirements to the terminal could be called essential and how the chosen requirements should be worded. Each country had expressed itself in its own way and was of the opinion that their wording of the re-quirements was the best one. In Denmark we had e.g. a requirement for a certain real part of the ringing impedance plus a certain numerical value of it. Germany had a re-quirement for a certain capacitive value of the impedance plus a numerical value. The requirement in TBR21 might e.g. be the common window valid for all countries – if such a window existed!<br>If it had been my baby I had insisted on first defining what was essential and then find-ing possible common windows on these points. Then it was not necessary to discuss those items any more.<br>A significant difference from NET4 was that TBR21 should only apply to a single ter-minal connected directly to the NTP (Network Terminating Point). The requirements in the Danish part of NET4 were based on having up to 200 Ω and 1,5 dB attenuation be-tween the NTP and the final terminal. Thus, the network will deliver at least 40 V AC as ringing voltage over the minimum ringing impedance connected in the NTP, while the requirement to the final terminal is that it works on 35 V AC ringing voltage. The re-quirements are also based on having more terminals in parallel, so the ringing imped-ance of each terminal must be a multiple of the minimum ringing impedance required by the network. All requirements in NET4 had to be adjusted in this way to apply in NTP for a single terminal before the common windows could be defined.<br>There would of course be several points left where it was not possible to find a common window. The discussions could then concentrate on these items.<br>I had arrived so late in the group that it would be impossible to introduce a new method for deciding the values. The efforts of us Danes was therefore only to see to it that TBR21 did not contain items which we could not comply with and for the rest try to help finding acceptable compromises when we had no special requirements to fight for. Typical was the French insistence on a requirement that the terminal should limit the loop current to 60 mA. All others had to accept this and worked then in unison to find out how the requirement should be worded for loop currents under 60 mA in such a way that it would still be possible to produce the terminals. The compromise found was that the terminal had to allow a loop current of at least 40 mA on the shortest loops.<br>Telecom invited on a trip one evening to a vineyard as it was in the period of wine har-vesting. We learned about the production of wine and it was a surprise for me that the first fermentation only takes about 14 days. During this time it is dangerous to enter the fermentation room due to the amount of carbon dioxide produced. The vineyard ran a restaurant in the village where we had ham and sausages – and an opportunity to taste their different wines!<br>Jørgen Lindegaard moves to GN Great Northern<br>In the autumn of 1995 it became impossible for Jørgen Lindegaard to continue as CEO of Copenhagen Telephones. His chemistry did not fit in with that of the CEO of the Tele Danmark concern, Hans Würtzen. Jørgen decided to leave Tele Danmark for the job as CEO of GN Great Northern, the oldest telecommunication company of Denmark, dating back to the 1850es when it was formed by C. F. Tietgen as a telegraph company.<br>A large farewell reception was held and I came of course too to say good by and good luck to Jørgen. He surprised me by not only accepting the good wishes but by asking me immediately whether I had a place in the new organisation due to start on January 1st, 1996, and which meant the dissociation of the old telephone administrations. Indeed I had, and he was glad to hear this. I had not expected him to think about other things than his own farewell to Copenhagen Telephones on this occasion, but was happily surprised.<br>Changes in the Telecommunication Supervisory Board<br>There was another change this autumn. In the TSB, Telecommunication Supervisory Board, Ernst Hansen retired when he became 60 years and there was a farewell recep-tion to wish him good luck.<br>A successor was called for via advertising and I must admit I felt myself qualified for this job. This was confirmed when Poul Tækker who I knew from i.a. the N4G-group phoned and asked if I had seen the advertisement? Indeed I had but I would not apply. I did not feel on the right shelf in the new organisation of Tele Danmark (I came further away from the exchanges), but had decided to stay. I did not say this to Poul Tækker but answered in stead that I did not expect them to choose a person of 62 years and soon 63 to replace Ernst Hansen who had left at 60!<br>The TSB found a replacement in Helene Jensen who came from the Civil Protection Agency. We had a fine cooperation when I became chairman for the TAFK-group, see below.<br>Legislation on Telecommunications<br>During these years a lot happened within the telecommunication field concerning liber-alisations of not only the terminals but also the whole infrastructure, aiming at having a fully free telecommunication market in the EC from January 1st, 1998 (except in the most backward countries). The EC issued directives and they should be followed by legislation in the EC countries. In Denmark the liberalisation was even hastened so the area was free already in the middle of 1996. However, the necessary legislation was only ready during the first half of 1997. It caused several contributions to journals about this topic. I found that one of them, in “IT Standard News”, deserved a comment and sent the following to the journal, where it was published early in 1996:<br>“A few comments to the article “The EC commission changes its view on the need for legislation concerning telecommunications” in IT Standard News, October 1996.<br>About the background Birger Jacobsen (BJ) writes that the philosophy behind the rules for terminals was the necessity “to protect the network itself against damage caused by the terminals” and that this has caused a comprehensive set of rules. The need for pro-tection itself has, however, not implied a lot of rules, see e.g. the draft regulations such as TBR15 for terminals on leased lines under the ONP-directive (Open Network Provi-sion) which may only be justified by this requirement. The many rules will only come into play due to the essential requirement for cooperation with the network to make, uphold, release etc. connections. But these rules are also not very comprehensive as they must be limited to the mentioned functions. Thus, no requirements can be made as to the correct function of supplementary services in connection with a type approval.<br>Within the present legislation BJ mentions the rules concerning voltages in the Low Voltage Directive (LVD) and that “the Terminal Directive (TD) had therefore to sup-plement the LVD as to other voltages”. This is regrettably not correct. The safety rules in the TD (91/263/EEC) may only be complied with through a reference to the LVD. This means that a terminal supplied from less than 75 Volts DC shall meet no safety rules due to the TD. Quite reasonable, neither users nor personnel from the telecom op-erators will be killed if such a voltage is leaked. However, there are some non-electrical requirements which are not covered and must not be covered, such as a requirement that the sound level in a telephone must not be so high that the hearing ability is spoiled. When the TD is revised special conditions within telecommunications ought to be al-lowed treated in regulations based on the TD, such that this requirement could be rein-stated.<br>During his treatment of the history BJ mentions as an example of damage to the network that a terminal “makes the whole network die”. This is a good illustration of the need for a definition of the content of the essential requirements. It is of course also damage when a part of the network dies, or if the terminal sends signals with such a power that they spill over to other connections. At this point the Danish TSB has by the way been more liberal than the EC, as it has removed the rules for terminals on digital leased lines. The network terminal from the operator – to which the user terminal is connected – protects the rest of the network, and if it is damaged it will not influence other connec-tions. With the introduction of ONP for leased lines rules for these terminals appear again, as the network terminal is perceived to be a (protected) part of the network.<br>BJ also writes that one might (maybe) conclude that “it is more important to keep the telecommunication network alive than to keep the users alive”, referring to the fact that safety requirements, EMC requirements and requirements to the network are only based on voluntary standards. BJ forgets that with the references in the LVD and the EMC-Directive and with the references in the ONP-Directive to what the operators shall de-liver, these voluntary standards are made obligatory.<br>Finally, one can to the four bullets about the proposal from the EC Commission add:<br>• If the requirements are limited to damage on people and network while the co-operation with the network is based on an agreement between user and operator one might open for a situation where a terminal can not be used on more net-works. Should a situation again be made in which each operator (Tele Danmark, Telia etc.) supplies its own terminal program together with the subscription?<br>• One can not obtain interoperability and harmonisation by legislation for the net-works only. The terminals must also comply with certain requirements in order to cooperate with the network. These requirements should be the same for all terminals for a certain service, independent of the network operator.<br>• It must remain essential that terminals on the fixed network (for both switched and leased lines) do not disturb other connections. To increase the level indis-criminately can not be tolerated.<br>• The terminals must thus continue to comply with certain requirements, but it can be better to base their access to the market on declarations from the producers about this as it is already done concerning safety and EMC.<br>• It is correct to maintain that all use of radio frequencies must be regulated.”<br>The TAFK-group<br>I became a member of this group when it was started in 1994 to revise the technical regulations for terminals for analogue leased lines.<br>As a basis we had the Publication 12/90 from the TSB, which had as an appendix the old Circular 12 from the tele-administrations. These administrations had also made a proposal for a revision, which had not been approved by the TSB.<br>The group was established according to the new procedures in the TSB, i.e. as a group formed by the TSB for which Tele Danmark provided chairman and secretary plus pos-sibly other members free of charge. Other organisations might also provide members.<br>Leo B. Rasmussen from Jutland became chairman and Arne Crown Rasmussen from Funen became secretary. Members at large were Ebbe Christensen from DELTA Elec-tronics Test, Ole Skjoldstrand from the Society of Electronic Factories and me.<br>The starting point was that ETSI was standardising some types of analogue leased lines under the ONP-LL Directive (Open Network Provision – Leased Lines) which would replace some of the types provided up to now by Tele Danmark, and which we had to provide (they belonged to the minimum set of leased lines along with digital 64 kbit/s and 2 Mbit/s lines which had to be offered throughout Europe). As to these types the new technical regulations for the terminals should just refer to the CTR’s to be issued by the EC Commission based on the TBR’s from ETSI (Common Technical Regulations, resp. Technical Basis for Regulation). I had been a part of the planning within ETSI to prepare these regulations as a member of PT9V in the autumn of 1991.<br>There were four types of analogue leased lines for which this applied viz. 2-wire and 4-wire lines of normal and special quality. There were, however, only two TBR’s, resp. TBR15 for 2-wire and TBR17 for 4-wire terminals. The difference in quality came in what the network operators should deliver (e.g. less variation of delay with frequency for lines of special quality).<br>But in addition to these types Tele Danmark also offered lines with signal conversion, with local quality and with broadcast quality. For these types of lines we had to go in detail with the requirements to the connection of terminals in the new technical regula-tions. At the same time we should adhere to the Terminal Directive from the EC, which said that it was only essential requirements which should be complied with in order to get a type approval for a terminal. Considering leased lines this meant that it was only possible to require such properties that the telecommunication networks were not harmed.<br>This caused many discussions in the group. My starting point was that there should be as wide limits for the terminals as possible, while Leo rather wanted more limitations. As to lines of local quality which were supplied as wire pairs between the customer ad-dresses (and were therefore normally limited to have both end points within the same exchange area) Tele Danmark specified that they were for transmission up to 15 kHz. In real life wire pairs can be used at much higher frequencies, e.g for ISDN2 connections using power up to about 50 kHz (and think of the DSL lines during the 2000es). My viewpoint was that customers should be allowed to use the lines at the highest frequen-cies they wanted, e.g. to an ISDN2 terminal under a PABX but at another address. It should only be required that they should at each frequency send at such a level that cross talk could not disturb neighbouring circuits. It should not be required to select wire pairs for this type of lines.<br>In the IEEE Communications Magazine there had been an article (Chen and Waring, May 1994) stating that the crosstalk attenuation decreased by 15 dB per decade, i.e. with 15 dB from 10 to 100 kHz and again with 15 dB from 100 kHz to 1 MHz. The labora-tory of Jutland Telephones could confirm that this was also the case in Denmark (it had been measured when ISDN was introduced). As a consequence this value was stated in the technical regulation as valid for the range from 15 kHz to 1 MHz.<br>In the spring of 1995 we could send our draft to the TSB and during the summer it was published for comments. The TSB had divided it in 3: TB95069 referring to TBR15, TB95070 referring to TBR17 and TB95071 for terminals on the special Danish lines. Comments were received from Tele Danmark to the latter and then nothing happened for a long time.<br>The reason was internal changes in the TSB. Ernst Hansen retired and Helene Jensen replaced him. She came from outside the TSB and had to work herself in first.<br>In the meanwhile other things happened. The EC Commission had decided that the only case in which the essential requirement for cooperation between terminals via the public network should be applied in connection with analogue terminals was for speech trans-mission between normal telephone sets directly connected to the NTP’s (Network Ter-minating Points). As a consequence the requirements to transmission and to the inter-face towards the extension, such that transmission from a normal telephone was possi-ble, in the technical regulations for PABXes were to be discarded. We had in TB95071 referred to exactly these requirements when writing the rules for terminals for lines with signal conversion from and to the usual telephone signals (ringing voltage and loop in-terruptions) on 2-wire lines.<br>Thus the group was again activated in the autumn of 1996, but in the meantime the two others from Tele Danmark had moved to other jobs within the company and could not continue. I had then to continue as both chairman and secretary. The work to include the requirements directly in TB95071 was, however, rather easy. We had only one meeting and managed for the rest with a few faxes before we could in the middle of November send our draft to the TSB.<br>This did not end our job. It would only be finished when both a Danish and an English version of the technical regulation had been issued. We took this up with the TSB, which were also very responsive to our proposal that the final editing of TB95071 should follow our proposals closer than was done in the issue published for comment.<br>It is lightly ironical that the requirements added directly to the technical regulations were for a disappearing circuit type. Tele Danmark could not any more buy the signal converters necessary between the physical line and the transmission equipment. At the same time the transmission equipment always bundled the circuits in groups of 30 cir-cuits (PCM systems) which had to follow each other. This was not economically sound if you only needed one circuit. The signal converters could be replaced by interfaces to the LLSN (Leased Lines Services Network) which was being introduced and which had the required 2-wire interfaces. But between such interfaces there was still the require-ment that 30 circuits had to be provided in parallel.<br>Tele Danmark had other cards in its sleeve. Digital public exchanges were introduced in great numbers and their facilities could be used. Especially the Centrex offering, a coun-try-wide PABX functionality with no other equipment at the customer address than normal telephone sets, was useful. Using this offering as the main PABX and with cus-tomers having PABXes and normal sets under it the same functions could be provided as with leased lines with signal conversion. In addition the customer should not – as with leased lines – pay whether he used the line or not (leaving it also to Tele Danmark to somehow cover the cost of 29 maybe unused circuits). Now the connections were in the switched network and existed only when needed. When a connection was not in ac-tual use the different parts of it were available for other users.<br>The Part 8 group<br>Part 8 of the Danish requirements to PABXes treated their access to “customer links” or 2 Mbit/s PCM connections to the public exchanges. These are structured PCM systems, meaning that the bit stream is divided in 32 timeslots of 8 bit each, repeated 8000 times per second. Timeslot 0 always has a certain pattern, making it possible for the receiver to recognise it and find all timeslots (and bit). Timeslot 1 to 15 and 17 to 31 are the 30 speech channels in which speech (or tone signals) are coded according to the A-law (in Europe). Timeslot 16 is the signalling channel, in this case using CAS or Channel-Associated Signalling, meaning that every 16th timeslot 16 has a certain recognisable pattern, making it possible to associate the 15 remaining timeslots 16 with 2 speech channels each. Thus each speech channel is accompanied by 4 signalling bit, repeated every 2 ms.<br>The signalling bit are used for line signalling (seizure, release, answer), while the regis-ter signalling (forward signals like dialled digits, type of caller, backwards signals to call for next digit or caller type or to report if the called party is free or busy) were MFC (Multi Frequency Code) signals in the speech channels. Already in 1974 I had worked with this, although on analogue lines, for the DDI (Direct Dialling In) function in EBX 8000 for CTBD and later for the state network of PABXes. From the beginning this sig-nalling was very flexible, as proved when it was discussed with Siemens in 1992. Unlike other PABX suppliers they wanted to receive the signal for type of caller before the dialled digits, but this was no problem, if only they asked for it with a backward signal and then asked for the relevant digits afterwards.<br>In 1990 the PABX-group had issued Part 8 in a “provisional draft” form. It could not remain so. The reason was that there did not exist at that time a real possibility for test-ing whether the terminals complied with the requirements. In 1994 the standardisation of ISDN was so far that it was at least possible to make proper test specifications for the physical connection.<br>Thus, the TSB created in 1994 a group besides the PABX-group to make a revision of Part 8. It did of course mostly consist of people from the PABX-group. Tage Fox Maule from TDBD, Tele Danmark Business Division, became secretary and should edit the draft.<br>While the group was active changes occurred in what should be required from PABXes. It was laid down that requirements should only apply to the connection to the line to the public exchange and be limited to the basic functions. Transmission requirements should be scrapped, as the essential requirement for cooperation between terminals in relevant cases, article 4(g) in the Terminal Directive, should only apply to normal tele-phone sets for live speech when the sets were connected directly to the analogue Net-work Terminating Points. The TSB was also becoming stricter as to what a Technical Regulation might contain. It should in the future be limited to the requirements and the test arrangements making it possible to verify the requirements.<br>The total set of PABX requirements was therefore cut down considerably. There was still a Main Regulation (the old part 1 and 2 with abbreviations etc.) but part 3 about analogue connections to the switched telephone network was replaced by the rules in NET4, edited by the N4G-group. Part 4 about analogue connections to leased lines ex-isted still in 1997 (when I retired) but would then be replaced by the rules edited by the TAFK-group. Parts 5 and 6 about supplementary equipment and functions on analogue connections to the switched telephone network were scrapped. Part 7 about the connec-tion of analogue telephone sets on the extension side was also scrapped. Part 8 about connection to customer links should remain and be edited by this group. Part 9 about connection to digital leased lines was replaced by a reference to the Technical Regula-tions for terminals on such lines. Parts 10 and 12 about transmission in resp. analogue and digital PABXes were scrapped. Part 11 did never exist.<br>As said: Part 8 should remain. And considering that there was no Technical Specifica-tion from Tele Danmark for its delivery of customer links we did for a long time neglect the new guide lines from the TSB and kept an informative chapter with such a specifica-tion. We did of course also advise that Tele Danmark should issue such a product speci-fication.<br>For the rest the job was a careful examination of the old draft and the solving of all doubtful points. A funny aspect was the scope of requirements. No test laboratories were interested in making a set-up for automatic testing of terminals for customer links, the market for tests would be too small for it to pay off.<br>The outcome of this was that when we handed over our result to the TSB in October 1995 there was an informative chapter and the requirements were limited to cover only the physical access, as it could be tested in the set-up already used for test of ISDN30 (or primary rate). Whether the signalling worked towards the public network we left to be based on suppliers’ declarations.<br>The TSB did not like this – they had no authority to implement such a rule. They did neither like the informative chapter.<br>In 1996 the old telephone administrations had been reorganised into functional entities in the Tele Danmark concern and the work with specifications was arranged in a new way. One point was that John Hartrup in NIS (Network Infrastructure Switching) should make a technical Specification for customer links. He later moved to the regional or-ganisation for Copenhagen and the work was taken up by Jürgen Endres in NUS (N Development S) with assistance from Harly Østergaard Jensen, U (Development). They were both members of the new PABX-group under the TSB. This resulted in TDK-TS 900234 of May 1997.<br>This meant that the informative chapter in the Technical Regulation could be replaced by a reference to this TDK-TS. Thus, in June 1997 the TSB could finally issue TB97087 with access requirements for terminals on customer links. I did not see it before my re-tirement, so I do not know the result of the considerations concerning the scope of test-ing.<br>The BELA-group<br>A group was formed within Copenhagen Telephones in 1995 to describe limitations for electronic solutions in the access network. The group should compare such solutions with having a copper pair to the customer.<br>Copper wires can do much more than transmit the necessary frequencies for speech in the range 300 to 3400 Hz. Their attenuation does not increase steeply above this band, in stead it increases gradually with the square root of the frequency. If the attenuation is 3 dB at 1000 Hz, it will be 6 dB at 4000 Hz etc. Therefore one can work with rather high frequencies such as ISDN2 connections requiring transmission up to about 50 kHz or even ADSL (Asymmetrical Digital Subscriber Line), using the copper pair to trans-mit TV-signals at several MHz to the customer.<br>An electronic solution might be used in a place where there were no more copper pairs to avoid laying a new cable, by e.g. installing electronics at the two ends of the fully used cable to double its capacity by transmitting ISDN2 signals on its pairs and translate these signals to normal copper wire signals at each end of the “narrow” cable. But this will limit the transmission for the connections to the speech band.<br>This is not the only limitation. The copper wire is also used to feed the telephone instal-lation such that it will work even if there is no electrical power in the house. It might still be possible to feed two installations via a single pair, but if more connections shall use the same pair a local power supply is necessary including a battery if the installation shall be independent of the electrical power system. However, when there is power via the copper pair one can also feed with one or the other polarity and this can provide ad-ditional signals to the installation, e.g. a polarity reversal when the called party answers. These signals or short interruptions of the feed current are impossible to use with most electronic solutions.<br>An important point is that customers subscribing to one or another signalling possibility shall not be transferred to electronics which can not transmit the signal. In order to avoid this one must at first map which limitations the different electronic solutions will imply. We choose in the group to make a list of the basic services, such as polarity reversal, and whether they could be delivered by the different electronic solutions. We then made a list of customer services and their need for the basic services. The sales staff needed such a list when they discussed with the customers.<br>The work led to a report, the last one before Copenhagen Telephones was disbanded at the end of 1995. Within the Network Division Johnny Damgaard continued the work and added in 1997 another pair of solutions and their limitations to the report.<br>The Housing-group<br>In the middle of 1996 all delivery of telecommunication services in Denmark was liber-ated. The laws were not quite ready, this happened only a year later, but it was obvious that there would be several requirements to Tele Danmark as it was the dominating sup-plier. One was the requirement that others could demand to get room in the buildings of Tele Danmark for their equipment to obtain “interconnect”, viz. the coupling between the networks.<br>It would be a matter of consequence that there should be rules for such interconnections. The other operator of mobile telephony, Sonofon, already had this type of interconnec-tions, otherwise their subscribers could only connect to each other, not with the users of the fixed network. In that case there would be only a few people preferring Sonofon, even if calls in their network were free!<br>The starting point might have been similar interconnects as those Tele Danmark had with telecommunication operators abroad, i.e. connections via transmission systems between equipment at the premises of each operator and with agreed protocols for transmission and signalling on the connection. But as stated the requirement here went further, the operators should have a right to install their equipment in the buildings of Tele Danmark.<br>Different groups were formed to consider the consequences for Tele Danmark and pro-pose rules for “interconnect”. I participated in a group evaluating the consequences of the requirement for collocation and we met repeatedly during 1996.<br>Our starting point was that interconnect should take place at a rather high level in the network, just like the connections to operators abroad. A number of exchanges through-out Denmark were chosen in which other operators could have their equipment installed and be connected to the Tele Danmark network.<br>A problem was and is that Tele Danmark has obligations towards the Danish state con-cerning the safeguarding of telecommunications under all circumstances. For this reason it can not be allowed for everybody to enter the buildings. This caused another require-ment to the buildings where interconnect should be possible: They had to be divisible in an area with access for other operators and an area exclusively for the Tele Danmark staff. Collocation was therefore divided in three classes, either each operator had his own locked room with a common access from the outside, or the operators had a com-mon room with access from the outside, or – as the third possibility – the equipment was placed in the same room as the equipment of Tele Danmark. In that case it was required that there were always personnel from Tele Danmark present during access.<br>Luckily there were already examples of other operators of especially data services hav-ing their equipment collocated with Tele Danmark equipment. The agreements about this was a good starting point.<br>However, collocation could not be limited to those few places in Denmark. A new op-erator might be a supplier of TV distribution who extended his plant to replace the Tele Danmark access network, i.e. the network between exchange and the single user. Such an operator needed access at a subscriber exchange and the rules had to adapt to this. In fact it applied not only to the need for collocation (where access to a transmission sys-tem in most cases could cover the needs) but also the subscription price. Tele Danmark had to realise that the large expenses in the access network disappeared in this case, such that the new operator could charge the fixed pay for the subscription. But there was compensation: The new operator was responsible for the whole access network and Tele Danmark should not any more make up bills to all subscribers, the new operator was their sole representative.<br>In the end of 1996 the work was halted as there was doubt about the legislation and its consequences for Tele Danmark. Early in 1997 the chairman of the group was delegated to Newtelco in Switzerland, which had been partly bought by Tele Danmark, and he was later employed by Newtelco. Therefore the group was inactive up to the time I retired.<br>The MCID-group<br>I became part of another group in 1996 and 1997, but not as a proper group member. The group should make proposals for a change of the alarm service such that it would not be necessary to have a specially developed program for the AXE exchanges from LME to run it. MCID means “Malicious Call IDentification”.<br>The reason for not being a proper member was that this matter applied to exchanges and my department, NUI, had nothing to do with them, they were the field of NUS. But con-sidering my past I knew about exchanges and signalling, and therefore the group wanted my active participation.<br>The specially developed program saw to it that when there was an alarm call the AXE exchange switched in a tone receiver in parallel on the line during the whole call. When the alarm centre pushed a code on their telephone set the receiver caused a signal to the supervision centre of Tele Danmark, where all exchanges are supervised at all times, printing out the number of the caller. The staff here could find the address of the caller and – when the alarm centre called – tell it.<br>The standard program in AXE for MCID did not switch in a tone receiver constantly. It was only called in (and sent a dial tone) when the called party (i.c. the alarm centre) sent an R-key signal, a short interruption of the current in the subscriber line. The develop-ment of the special program for Denmark had a higher and higher cost for each new program package for AXE, so it would be an economical advantage to use the standard program.<br>But this was not only about an extra push of a button in the alarm centre, a push already introduced in several places in Denmark. Many alarm centres used a PABX, possibly connected via a 2 Mbit/s link to the public exchange. Thus, the real problem was whether the used PABXes could send the R-key signal to the public exchange? An R-key signal was already used in PABXes (from long ago) to call for a new dial tone to start a call-back and transfer. The PABX would thus have to, after having received an R-key signal, perceive a special dialled code as a message meaning that it should send an R-key signal to the public exchange. There would then come a dial tone from this exchange and the code for printing the calling number could be dialled. The user at the alarm centre should in this way both push the R-key and dial the special code in addition to what he was used to. This was not very ingenious!<br>In addition to this all alarm calls in Copenhagen go to the fire department, who transfers the call to the police if it is about other than fire or ambulance services. In this way the police might have to send three codes after the R-key signal before the last one gave a print-out of the calling number!<br>One of the solutions considered was to make all connections to alarm centres into ISDN connections. In this case there would at all times be a special signalling channel avail-able independent of the speech channel to send the necessary signal from a function key during the whole call. However, such a change was not cheap and might require new PABXes in the alarm centres.<br>Finally, the solution could not just be a proposal from the group introduced over night. It had to be field-tested before coming in actual use. This had to wait for the next pro-gram package for the AXE exchanges to test in connection with the standard program. Therefore the group had not finished its job when I – along with several others – left Tele Danmark.<br>Retirement<br>In the summer of 1997 the message came from the CEO of Tele Danmark: All employ-ees born in 1940 or earlier could terminate their employment at the same conditions as if they had been fired. This suited me fine, I would shortly be 65 years of age, the normal pension time if I had stayed with Philips, and the economy was OK. Further, like all oldies I could expect that if I did not resign, I would be fired! Thus, by the end of Au-gust 1997 my professional career ended...
+
<br>In 1992 the N4H group was started with Knud Solholt from Jutland Telephones as chairman. He was also the Danish representative in the ETSI technical committee TE5. This committee was responsible for ETS 300 001 or NET4, the rules for terminals on the analogue switched telephone network (or PSTN, Public Switched Telephone Network). NET4 from 1991 was a catalogue over the different rules in Europe, as national deviations were OK then. The differences were only removed when a change was made to TBRs (Technical Basis for Regulation), and TE5 had started a project to harmonise these rules. I became part of the N4H group because I had been part of the ETSI project team PT26, NET4 editing group, in the autumn of 1990.<br>The N4H group should evaluate the papers from TE5 from a Danish viewpoint. The name meant of course NET4 Harmonisation. The papers started with a discussion of which requirements to the terminals it was possible to harmonise and then went on to what the harmonised values should be.<br>We took most of the European discussion very quietly. We let the big network operators carry the burden and only occasionally our discussions made Solholt try his influence in TE5.<br>One of the essential questions was whether the loop resistance could be harmonised? The loop resistance is the DC resistance a terminal in its off-hook state presents to the network, such that the network perceives the terminal as busy. The Danish requirement in NET4 was that below 16 mA loop current the voltage between the terminal connec-tion points should be below 6,4 V, and above 16 mA the DC resistance between the connection points had to be less than 400 Ohm. This value or a lower one could very well be harmonised in our opinion. Although it might be a tightening related to what was required in Denmark if it was reduced to e.g. 300 Ohm, there would be no problems getting terminals complying with the requirement if it was a harmonised value.<br>However, an item quite impossible to harmonise was the French requirement that a terminal had to limit the loop current to a maximum of 60 mA. They stated it was necessary as&nbsp;a higher&nbsp;current&nbsp;from the line circuits in the exchange would saturate them and cause a lower quality of the transmission if the current was higher. Everybody else found that this was a problem handled by the exchanges themselves, making the terminals responsible could only be considered a technical trade barrier.<br>The discussion in ETSI continued for a long time after our group had changed into the N4G group with the task to revise the Danish part of NET4. Only in November 1994 the ETR 075 was issued with the result of the efforts by ETSI to harmonise NET4.<br>
 +
 
 +
=== The DECT group ===
 +
 
 +
<br>Another group in which I participated from 1992 was the DECT group. DECT is Digital European Cordless Telephony and is a technique to allow many users to have each his cordless telephone network in use within a small area. DECT is not an independent network like the mobile telephone network, each DECT system works under a public or private (PABX) exchange or network of exchanges in the PSTN. It could be used e.g. by the owner of a PABX to give its extensions local mobility unhampered by a neighbour giving the same opportunity to his employees.<br>LME (L. M. Ericsson) had done early work in this area and had developed and had in test operation several systems according to the same principles, but at frequencies around 900 MHz. The standards were now so far that a range of frequencies around 1900 MHz had been reserved and various protocol questions had been settled. LME could supply systems according to the latest version of the standards and the Danish telecom administrations would like to test them in different applications.<br>An arrangement was made for equipment for four tests, one at each administration. In Jutland it was set up under a public exchange in Aalborg as an alternative to the copper pairs in the subscriber cables, In Southern Jutland it was set up under PABXes in the frontier town Padborg to give local mobility to the personnel of the transportation firms. On Funen it was set up under the PABX in the headquarters of the administration to give mobility to the employees. In the Copenhagen area it was set up in a shopping centre in Hillerød under the public exchange. The handsets should work in parallel with people’s PSTN telephones, such that the users could have their telephones with them when they were shopping.<br>The DECT group prepared the tests and should evaluate the result. It was indeed started but died out due to the changes in the organisation (i.a. the fusion of the business divisions). It is my impression that there were never many customers to the application in Hillerød. Maybe because most people do not want to be called on the telephone when they are shopping – and the others use a proper mobile telephone, the use of which was not limited to one shopping centre!<br>
 +
 
 +
=== Øberg leaves CTBD ===
 +
 
 +
<br>Already before the fusion of the business divisions the “war” between them had caused heavy discussions within each administration. This also led to Øberg taking his leave. Some years later when I met him in Horsens he said it was a disagreement about what CTBD should do between him and the chairman of the board, the CEO of CT Jørgen Lindegaard, which had caused him to quit his job.<br>He was not unemployed for long. He told me that he was called by the CEO of Alcatel Kirk with an offer to replace him. It suited him better to become CEO of S. C. Sørensen in Randers (I think it was a family business). Thus Øberg was employed here and travelled between their sites in Ballerup and Horsens until the Alcatel operation in Horsens was closed down. Their PABX production could not any more compete with the large players on the market and Alcatel could in stead import PABXes to the Danish market.<br>Other parts of the production of Alcatel could also not continue, but they were bought by the employees and continued under the old name Kirk. It was the production and further development of transducers, i.e. microphone and telephone capsules for fixed and mobile telephones, a large export item as the employees were and are experts in this area. And it was the production and further development of telephone sets sold both in Denmark and abroad. Alcatel concentrated on the production of space electronics in Ballerup and the sales of telecom equipment from abroad.<br>I also left the fused TDBD, but not on my own initiative and far less dramatically. Small PABXes up to 40 extensions should now be sold by the main companies, so many employees from TDBD were transferred to CT and Jutland Telephones. I was one of these although I never had worked with such small PABXes while in CTBD. I came to the CT Telecom Division in the spring of 1993.<br>
 +
 
 +
=== ETSI STC BTC2 ===
 +
 
 +
<br>By January 1st 1992 those of the employees of CTBD who had from old times been em-ployed in line with public servants, should decide whether they would be employed on a contract in CTBD or return to “big CT”. I.a. Niels Kristiansen, who was in the PABX group, chose to return. He then had to leave the PABX group and I proposed – and got approved – that he was replaced by another CTBD employee, Torben Paaske.<br>Niels also had to leave the work in ETSI STC BTC2 (or BT2 as it was called then) where he participated along with Jørgen Hess Kristensen from JTBD, Jutland Telephone Business Division, (who was the Danish spokesman). I suggested that I should replace Niels and this was accepted. I participated in meetings in Munich May 1992, in Bath October 1992, Sophia Antipolis April 1993 and Oslo September 1993.<br>Munich is the home town of Siemens and CTBD had started discussions about the in-troduction of HICOM in Denmark with them a few months earlier. This should be kept a secret for our competitors in JTBD, so I kept mum towards Jørgen. Well, it corre-sponded very much to what I was used to from Philips. We were always very friendly towards people from our competition but kept mum about company secrets. And so did the others!<br>When I received the flight ticket, I also got a telecard (to use in pay phones) with a value of DEM (Deutsche Mark) 5 from Lufthansa. It had a Danish text urging to use the card during the trip. I did so, but during the walk to the meeting place I passed a shop on a corner of the main station advertising the buying and selling of telecards. This I would of course test and every price from DEM 1 upwards would be a profit! Thus, when the card had been used I visited the shop, the card was set in a reader (to check the remain-ing charge and to ensure it was a real telecard) while I muttered that a German telecard with Danish text had to be a rarity. Well, they offered 2 DEM for it which I gladly re-ceived.<br>The German Telecom, who had arranged the meeting, invited all participants on an eve-ning trip to Garmisch-Partenkirchen, where we had a good meal and enjoyed local per-formers – and a performance of some of us, who showed their muscle power by compet-ing about who could hold a filled beer glass in an outstretched hand for the longest time.<br>The meeting in Bath was called by British Telecom. I used the opportunity to postpone my trip home to Saturday, such that I could be a tourist Friday afternoon. Going by bus around in the town, visiting the Roman Baths and the Cathedral, the latter with i.a. the epitaph from one great actor to the other, de Garrick writing over James Quin with the potent final words: “Whate,er thy strength of body, force or thought; In nature’s happi-est mold however cast; To this complexion thou must come at last”.<br>Several years later I found these names in the book by Knud H. Thomsen “Røverne I Skotland” (The Scottish Robbers). But back in 1992 I had borrowed the book “Sarum” at the public library. This book is about Salisbury from the Ice Age to the present. This town and Stonehenge are not far from Bath and Bath is mentioned at about the time when the Roman Baths were built.<br>At the meeting in Oslo I was accompanied by my wife, Jytte. We made it a small holi-day, first a couple of days in the Swedish archipelago north of Gothenburg and after the meeting a Friday afternoon as tourists (which my wife had also been every day during the meetings). The meeting was well attended from Denmark, with 4 participants, i.a. Bent Qvist from Copenhagen Telephones just like me. This was because the last items in the standard for analogue leased lines should be decided at the meeting.<br>Friday Jytte and I made two trips to the woods North of Oslo. First we went by rail to a place with a beautiful lake which we circled, then by another railway to the Frogner hills. We made again a long walk and enjoyed then hot chocolate in a restaurant with a view to Oslo where we saw the boat to Copenhagen sail out.<br>Spring 1993, the Rouban Group<br>During the meeting of BTC2 in April 1993 I became chairman for a small sub-sub-group.<br>BTC2 was busy with standards for leased lines, the task I had helped planning in the autumn of 1991. The job was done in project team PT22V and it was then discussed in working party WPF of BTC2. My place was in WPA on transmission, but I got mixed up in the plenary discussions of the results of WPF. In any way the requirements to a terminal on an analogue leased line as to the power it might transmit outside the speech band was thoroughly discussed at the meeting in Sophia Antipolis.<br>The problem was that while the requirements to the connection, which the telecom ad-ministrations should comply with, were comprehensive in order to ensure that the cus-tomers got a connection of a certain quality (as regards attenuation, distortion, delay, noise etc.), the requirements to the terminals were governed by the Terminal Directive. It lists 7 essential requirements to which a terminal must comply, viz. 1) user safety, 2) safety for the telecom personnel, 3) EMC (no disturbances to and from other apparatus), 4) no harm to the network, 5) a reasonable use of radio frequencies, 6) cooperation with the network to make, keep and release connections and 7) cooperation with other termi-nals via the network in justified cases. Requirements 1), 2) and 3) shall be complied with through other standards and 5), 6) and 7) were irrelevant for terminals on leased lines (no radio transmission, the connection exists all the time and can not be changed from the terminal and 7) was to be considered for speech transmission only). This left only 4), no harm to the network, and what did that mean?<br>There were fanatics (“liberals”) who wanted to limit this to mean only physical damage, such as when a high voltage from a terminal could make network equipment fail. We obtained however agreement that it should also mean that a terminal was not allowed to send signals disturbing other users of the telecom networks. Such was the stance in April 1993. I had made a proposal ensuring that the signals sent could not via crosstalk (i.e. coupling to other wire pairs in the cable) disturb others. This was however not enough. During the meeting it was emphasised especially by the participants from Brit-ish Telecom that transmission systems where connections are multiplexed on a common path did not isolate the channels completely from each other. On a carrier system the channels are placed besides each other within a range of frequencies. If a channel sends too much power at the frequencies of a neighbouring channel it will disturb. The filters for each channel are not ideal, partly as a question of price, partly as a question of the technique. If a filter cuts off unnecessary frequencies too sharply it will distort the wanted frequencies close to the edge. All this was a quite justified viewpoint.<br>The discussion made the Chairman of BTC2, John Horrocks from British Telecom, pro-posed a group formed to present a proposal for requirements at the WPF meeting in July and he asked me at the meeting if I would chair this group.<br>I made a long pause. Yes, I would like to, but I had recently come to CTTD, Copenha-gen Telephones Telecom Division, (when CTBD and JTBD were joined into TDBD and small PABXes were moved to CTTD), and I had sensed the attitude to the work in ETSI. Thus it was clear for me that the job as chairman had to be handled without trav-elling. It was also only for a short time. The thoughts ran through my head until I saw a way and accepted the job. I collected names and fax numbers of the interested BTC2 members.<br>As soon as I was back home I described my ideas in a fax to the interested people. At the same time I coined the name The Rouban Group for “Requirements to OUt-BAnd Noise”. The idea was to send my proposal as far as it went by fax on May 1st, June 1st and July 1st, and then the interested parties could fax their comments by the 15th of each month. The schedule was to call for proposals in May, to make a draft by June and a final draft in July. There would just be time to get a yes or no to this before the WPF meeting in the end of July.<br>It was carried out as proposed. There was a good proposal from Holland and it was re-written to the standard formulation of ETSI. There were some comments but nothing disastrous in June, so it was largely this proposal which went out in July. It was unani-mously accepted (at least nobody had comments to the changes from June to July). Brit-ish Telecom even wrote that they were very satisfied and could support the proposal 100%. It also passed through the WPF meeting without amendments except for a few minor adjustments. One was that L. M. Ericsson rather late had observed that the im-pedance over which the power was measured was not the correct one outside the speech band. I had confirmed this in a comment to the final proposal and the change was intro-duced.<br>The problem was that inside the speech band a fixed impedance is used equal to the im-pedance at 1020 Hz, although the impedance changes over the band. When one moves outside the speech band and as here even up to 1 MHz the impedance deviates however too much from the impedance at 1020 Hz to keep it at this value.<br>The results from the Rouban-group were included in the requirements to terminals for analogue leased lines according to the ONP-LL directive, TBR15 and TBR17.<br>It is my impression that much standardisation work could be done as in the Rouban-group by contact via fax or e-mail, when people have first met and got to know each other. It is also my impression that the few meetings are well prepared by these means so everybody is fully oriented about all viewpoints before them.<br>The N4G-group<br>During 1993 the work on NET4 continued in the N4G-group. The job was to revise the Danish part of NET4 (or ETS 300001) to be part of a new version of the standard.<br>The first version had an exception for PABXes and there was in stead a reference to national regulations for them. These words would be removed, but NET4 had on several points values different from those in the Danish requirements to PABXes.<br>The PABX-group was by the way busy revising part 3 of these technical requirements, which treated their connection to analogue subscriber lines in the switched telephone network. I was editor on this task. We were ready a couple of months before the N4G-group was ready, but the Telecommunication Supervisory Board decided that the new version should not be issued while waiting for NET4. All these efforts for nothing, but the work was of course valuable for the work in the N4G-group.<br>One of the great differences of opinion within the group related to the leakage resistance of the terminal to earth. Circular 2 (requirements of the network to subscriber terminals in general on the switched network) said that the subscriber installation as a whole was allowed to have a minimum of 10 MΩ from each wire to earth. The first issue of NET4 required a minimum of 100 MΩ from each wire to earth for each terminal (several might be connected in parallel). At an early moment we had seen in the PABX-group that this was a too small value for a PABX when it should measure the polarity over the wires (used to check if the subscriber line was busy in the public exchange or not, in order to release the line as soon as an external caller had gone on hook). The PABX would require at least 10 MΩ, and as there might be more terminals in parallel con-nected to the NTP (Network Terminating Point) Circular 2 had to allow a correspond-ingly smaller leakage resistance of the total subscriber installation. Considering that the requirement to the subscriber cable was a minimum of 100kΩ from each wire to earth or 100 times less than required in NET4 of the subscriber terminal there was an ample margin to exploit. We had already in 1989 discussed this point with the technical com-mittee of the telephone administrations and received an accept of a minimum value in Circular 2 for the whole subscriber installation of 3 MΩ from each wire to earth. We PABX specialists used this value repeatedly in the discussions in the N4G-group, as this was our basis for requiring the value in NET4 changed from 100 to 10 MΩ, such that NET4 could also apply to PABXes.<br>We did not succeed completely. The revised text measures leakage resistance between short-circuited line terminal points, i.e. in the worst case half as much as from each point to earth, and this value was set to 10 MΩ in general. A considerable reduction compared with the first version where the same value would have been 50 MΩ. How-ever, the text was made acceptable due to an exception for PABXes. They were allowed a leakage resistance between the short-circuited terminal points and earth of minimum 5 MΩ.<br>This implied of course that the term PABX had to be defined in NET4. NET4 already had requirements to terminals connected in series, defined as terminals with one port towards the network and one towards a NET4 terminal, where the second terminal only receives power for its operation from the first terminal. We worked much on the defini-tion of a PABX and got as the result that a PABX was any other terminal with these two types of ports. Thus, if there were only two ports it was a requirement that the terminal on the second port was at least partly fed locally and if there were more than two ports it was by definition always a PABX.<br>The work was ready in the end of 1993 so it could be notified within the EC and could be sent to ETSI. It was, however, not included in the April 1994 version of NET4. It was only included in 1995 and the Telecommunication Supervisory Board issued a Dan-ish version of NET4, which only included the common text with the Danish values and the special Danish amendments such as the definition of a PABX.<br>The PANT-group<br>In 1994 I became part of a group preparing a buyer’s specification for an NT, Network Terminal, for ISDN PRA (Integrated Services Digital Network, Primary Rate Access), i.e. the unit installed by Tele Danmark (TDC) at customer premises to provide the inter-face to a 30 channel connection on a 2 Mbit/s subscriber line. The specification should of course be based on the standards from ETSI, and they covered the digital line, errors on it as well as the T interface, which is the interface to the customer’s equipment. The group was established by IU-A, the Infrastructure Committee, Access, of TDC.<br>But as always there were options, possibilities for choice, in the standards and our task was to choose which of these should apply to the NTs bought by TDC.<br>Chairman of the group was Ole Philbert from Jutland. Members were Erik Bisgaard and Poul Erik Pedersen from Jutland and Otto Kimer and me from Copenhagen.<br>The purpose of such a specification is of course to make common rules for the behav-iour of the equipment, such that we were not tied to use only NTs from LME under LME exchanges, or NTs from Siemens under the exchanges they had supplied. It should also be possible to buy NTs from a third supplier and use them under both types of ex-changes. The more the different network elements can be independently bought, the better prices can be obtained by TDC!<br>We had many meetings discussing the possible choices and their consequences. With my background I insisted from the start that the NT we should specify ought to be us-able also on “subscriber links”, 2 Mbit/s digital lines with switched telephony used as subscriber lines to digital PABXes, and on 2 Mbit/s structured leased lines, which we should provide as a part of our obligations according to the EC directive about leased lines (ONP-LL) and where I had participated in the planning of the standardisation work in ETSI PT9V during the autumn 1991. It turned out that the various requirements fitted together, such that this could be realised.<br>It also turned out that I got the job of collecting the contributions and reformulate them to the required application specifications to the ETSI standards. There were three stan-dards, one for general conditions (ambience etc.), one for the T-interface and one for the digital line.<br>The job was finished in January 1994 and the group was disbanded a couple of months later. In July 1994 TDC then issued their technical specification TDK-TS 900213 with our specifications.<br>I got mixed up in the use of the specifications later. Poul Erik sent me compliance statements, i.e. confirmations that the requirements were complied with, from LME and Siemens for evaluation. I remember that Siemens did not comply with one of the re-quirements, that the NT towards the customer should set the bits S(a5) and S(a6) used on the line to the value 1. Siemens had chosen to set them at 0. Back to the ETSI speci-fication for the T-interface: it required that the terminal should ignore these bit. We had made the requirement mandatory, but could accept the choice made by Siemens. If we later had to revise the specification we would probably drop our requirement.<br>Later while in the network division (after the old administrations had been replaced by functional divisions in TDC) my boss, Martin Gregersen, came with a data sheet for such an NT from Fujitsu. I said to him that he should ask Fujitsu to make a compliance statement according to TDK-TS 900213 as a basis for further discussions. I do not know whether this led to our buying these NTs.<br>Finally, in 1997, Poul Erik asked me if I still had the files from the past. They had been included in INFR, ISDN Functional Requirements, and these should now become TDK-TSes. Thus it would be good to have the original files. Luckily I had always made back-ups on floppies and had brought them with me from CTBD via CTTD to the network division. Thus I could go back in spite of the change of PC and operating system and send them to him. Another piece of luck was of course that the floppies could still be read although I had changed PC!<br>The PSTN-group<br>In 1994 I became part of a group specifying the monopoly interface in the switched ana-logue telephone network. This was probably another consequence of my work in ETSI during the autumn of 1990.<br>Our chairman was Erik Bisgaard from Jutland. The group was established by IU-A, the Infrastructure Committee – Access network.<br>The monopoly interface or - as it was later called - the NTP, Network Terminating Point, is the point where the TDC cable meets the subscriber installation. The liberalisa-tion meant that not only were the terminals, but also the whole installation at the sub-scriber premises, outside the TDC monopoly in making telephony connections. A de-scription was required of what TDC delivered in the NTP and of what we delivered be-tween two NTPs, corresponding to the network interface and the connection properties in the specifications for leased lines.<br>We had a good starting point in the old Circular 2 (from the 1970es), which had a sec-tion dealing with the subscriber installation. There was i.a. a requirement for a minimum loop current of 15,7 mA when the NTP was terminated in 600 Ω, and the distribution of the 600 Ω was described (400 Ω in the final terminal, 200 Ω in the subscriber installa-tion including serially connected terminals). This had to be modified as the requirements to terminals were changed since the 1970es (the old Circular 27), e.g. the requirement to the final terminal at loop currents below 16 mA was not any longer a maximum of 400 Ω, but a maximum of 6,4 V over its terminating points, and this had to be reflected in the properties of the NTP. I emphasised that as far as possible we should specify the requirements to be in line with Circular 2.<br>At the same time IU-T, the Infrastructure Committee – Transport Network, was busy on a similar job. They rather looked at how little TDC could provide in the NTP and would base their efforts at the work in ETSI to make a common European set of requirements to telecommunication terminals in TBR21, i.e. requirements ensuring that a terminal might be used throughout Europe. Their attitude was that the network should not prom-ise better properties than just required to serve a TBR21 terminal.<br>My standpoint – to which my group adhered – was that it was not the network which had to adapt to TBR21, but the TBR21 terminal which had to adapt to the network. Cir-cular 2 was the Danish starting point. It said e.g. 600 Ω terminal resistance in the NTP. If another network operator required 500 Ω terminal resistance and this became part of TBR21 it would be OK for us, but should not change our specifications for the network. If on the other hand it was sufficient for all other European network operators to require a loop resistance of 800 Ω, we should require of TBR21 that its requirement was our 600 Ω. Thus, our group insisted on maintaining the 600 Ω loop resistance over the NTP, while the other group would adhere to the requirement of maximum 500 Ω loop resis-tance in the terminal according to TBR21, and then we should not promise that it would be perceived as a loop (off hook) if there were more than 500 Ω loop resistance over the NTP.<br>This ended up, as one can guess, with two specifications with slightly different contents. The shape of our draft became the final one, but it was decided in IU that the figures from IU-T should be used, which they were. I got angry during this, when it was re-ported that the chairman of IU-T, Ole Jahn Svinding from Jutland, had said in an IU meeting that there was agreement about the figures from IU-T. This was blatantly incor-rect. But in the final discussions Erik Bisgaard was under a large pressure, and therefore the figures from IU-T were included in the final specification, issued as TDK-TS 900221.<br>Of course I could not help laughing later when the figures in TBR21 were modified (and TBR21 was by the way not finally approved as late as in the spring of 1997), meaning that there is no more a reference for the figures in TDK-TS 900221!<br>ETSI STC TE5<br>A consequence of the work in the N4G-group and the PSTN-group was that I also be-came part of ETSI STC TE5. However, I took only part in one meeting, in Mainz in October 1994, invited by the German Telecom.<br>At this time TE5 was busy finishing TBR21, a common European NET4. It applied to non-voice equipment, a corresponding TBR37 should be issued for telephone sets with TBR38 valid for the speech transmission itself.<br>There had already been lengthy discussions of which requirements to the terminal could be called essential and how the chosen requirements should be worded. Each country had expressed itself in its own way and was of the opinion that their wording of the re-quirements was the best one. In Denmark we had e.g. a requirement for a certain real part of the ringing impedance plus a certain numerical value of it. Germany had a re-quirement for a certain capacitive value of the impedance plus a numerical value. The requirement in TBR21 might e.g. be the common window valid for all countries – if such a window existed!<br>If it had been my baby I had insisted on first defining what was essential and then find-ing possible common windows on these points. Then it was not necessary to discuss those items any more.<br>A significant difference from NET4 was that TBR21 should only apply to a single ter-minal connected directly to the NTP (Network Terminating Point). The requirements in the Danish part of NET4 were based on having up to 200 Ω and 1,5 dB attenuation be-tween the NTP and the final terminal. Thus, the network will deliver at least 40 V AC as ringing voltage over the minimum ringing impedance connected in the NTP, while the requirement to the final terminal is that it works on 35 V AC ringing voltage. The re-quirements are also based on having more terminals in parallel, so the ringing imped-ance of each terminal must be a multiple of the minimum ringing impedance required by the network. All requirements in NET4 had to be adjusted in this way to apply in NTP for a single terminal before the common windows could be defined.<br>There would of course be several points left where it was not possible to find a common window. The discussions could then concentrate on these items.<br>I had arrived so late in the group that it would be impossible to introduce a new method for deciding the values. The efforts of us Danes was therefore only to see to it that TBR21 did not contain items which we could not comply with and for the rest try to help finding acceptable compromises when we had no special requirements to fight for. Typical was the French insistence on a requirement that the terminal should limit the loop current to 60 mA. All others had to accept this and worked then in unison to find out how the requirement should be worded for loop currents under 60 mA in such a way that it would still be possible to produce the terminals. The compromise found was that the terminal had to allow a loop current of at least 40 mA on the shortest loops.<br>Telecom invited on a trip one evening to a vineyard as it was in the period of wine har-vesting. We learned about the production of wine and it was a surprise for me that the first fermentation only takes about 14 days. During this time it is dangerous to enter the fermentation room due to the amount of carbon dioxide produced. The vineyard ran a restaurant in the village where we had ham and sausages – and an opportunity to taste their different wines!<br>Jørgen Lindegaard moves to GN Great Northern<br>In the autumn of 1995 it became impossible for Jørgen Lindegaard to continue as CEO of Copenhagen Telephones. His chemistry did not fit in with that of the CEO of the Tele Danmark concern, Hans Würtzen. Jørgen decided to leave Tele Danmark for the job as CEO of GN Great Northern, the oldest telecommunication company of Denmark, dating back to the 1850es when it was formed by C. F. Tietgen as a telegraph company.<br>A large farewell reception was held and I came of course too to say good by and good luck to Jørgen. He surprised me by not only accepting the good wishes but by asking me immediately whether I had a place in the new organisation due to start on January 1st, 1996, and which meant the dissociation of the old telephone administrations. Indeed I had, and he was glad to hear this. I had not expected him to think about other things than his own farewell to Copenhagen Telephones on this occasion, but was happily surprised.<br>Changes in the Telecommunication Supervisory Board<br>There was another change this autumn. In the TSB, Telecommunication Supervisory Board, Ernst Hansen retired when he became 60 years and there was a farewell recep-tion to wish him good luck.<br>A successor was called for via advertising and I must admit I felt myself qualified for this job. This was confirmed when Poul Tækker who I knew from i.a. the N4G-group phoned and asked if I had seen the advertisement? Indeed I had but I would not apply. I did not feel on the right shelf in the new organisation of Tele Danmark (I came further away from the exchanges), but had decided to stay. I did not say this to Poul Tækker but answered in stead that I did not expect them to choose a person of 62 years and soon 63 to replace Ernst Hansen who had left at 60!<br>The TSB found a replacement in Helene Jensen who came from the Civil Protection Agency. We had a fine cooperation when I became chairman for the TAFK-group, see below.<br>Legislation on Telecommunications<br>During these years a lot happened within the telecommunication field concerning liber-alisations of not only the terminals but also the whole infrastructure, aiming at having a fully free telecommunication market in the EC from January 1st, 1998 (except in the most backward countries). The EC issued directives and they should be followed by legislation in the EC countries. In Denmark the liberalisation was even hastened so the area was free already in the middle of 1996. However, the necessary legislation was only ready during the first half of 1997. It caused several contributions to journals about this topic. I found that one of them, in “IT Standard News”, deserved a comment and sent the following to the journal, where it was published early in 1996:<br>“A few comments to the article “The EC commission changes its view on the need for legislation concerning telecommunications” in IT Standard News, October 1996.<br>About the background Birger Jacobsen (BJ) writes that the philosophy behind the rules for terminals was the necessity “to protect the network itself against damage caused by the terminals” and that this has caused a comprehensive set of rules. The need for pro-tection itself has, however, not implied a lot of rules, see e.g. the draft regulations such as TBR15 for terminals on leased lines under the ONP-directive (Open Network Provi-sion) which may only be justified by this requirement. The many rules will only come into play due to the essential requirement for cooperation with the network to make, uphold, release etc. connections. But these rules are also not very comprehensive as they must be limited to the mentioned functions. Thus, no requirements can be made as to the correct function of supplementary services in connection with a type approval.<br>Within the present legislation BJ mentions the rules concerning voltages in the Low Voltage Directive (LVD) and that “the Terminal Directive (TD) had therefore to sup-plement the LVD as to other voltages”. This is regrettably not correct. The safety rules in the TD (91/263/EEC) may only be complied with through a reference to the LVD. This means that a terminal supplied from less than 75 Volts DC shall meet no safety rules due to the TD. Quite reasonable, neither users nor personnel from the telecom op-erators will be killed if such a voltage is leaked. However, there are some non-electrical requirements which are not covered and must not be covered, such as a requirement that the sound level in a telephone must not be so high that the hearing ability is spoiled. When the TD is revised special conditions within telecommunications ought to be al-lowed treated in regulations based on the TD, such that this requirement could be rein-stated.<br>During his treatment of the history BJ mentions as an example of damage to the network that a terminal “makes the whole network die”. This is a good illustration of the need for a definition of the content of the essential requirements. It is of course also damage when a part of the network dies, or if the terminal sends signals with such a power that they spill over to other connections. At this point the Danish TSB has by the way been more liberal than the EC, as it has removed the rules for terminals on digital leased lines. The network terminal from the operator – to which the user terminal is connected – protects the rest of the network, and if it is damaged it will not influence other connec-tions. With the introduction of ONP for leased lines rules for these terminals appear again, as the network terminal is perceived to be a (protected) part of the network.<br>BJ also writes that one might (maybe) conclude that “it is more important to keep the telecommunication network alive than to keep the users alive”, referring to the fact that safety requirements, EMC requirements and requirements to the network are only based on voluntary standards. BJ forgets that with the references in the LVD and the EMC-Directive and with the references in the ONP-Directive to what the operators shall de-liver, these voluntary standards are made obligatory.<br>Finally, one can to the four bullets about the proposal from the EC Commission add:<br>• If the requirements are limited to damage on people and network while the co-operation with the network is based on an agreement between user and operator one might open for a situation where a terminal can not be used on more net-works. Should a situation again be made in which each operator (Tele Danmark, Telia etc.) supplies its own terminal program together with the subscription?<br>• One can not obtain interoperability and harmonisation by legislation for the net-works only. The terminals must also comply with certain requirements in order to cooperate with the network. These requirements should be the same for all terminals for a certain service, independent of the network operator.<br>• It must remain essential that terminals on the fixed network (for both switched and leased lines) do not disturb other connections. To increase the level indis-criminately can not be tolerated.<br>• The terminals must thus continue to comply with certain requirements, but it can be better to base their access to the market on declarations from the producers about this as it is already done concerning safety and EMC.<br>• It is correct to maintain that all use of radio frequencies must be regulated.”<br>The TAFK-group<br>I became a member of this group when it was started in 1994 to revise the technical regulations for terminals for analogue leased lines.<br>As a basis we had the Publication 12/90 from the TSB, which had as an appendix the old Circular 12 from the tele-administrations. These administrations had also made a proposal for a revision, which had not been approved by the TSB.<br>The group was established according to the new procedures in the TSB, i.e. as a group formed by the TSB for which Tele Danmark provided chairman and secretary plus pos-sibly other members free of charge. Other organisations might also provide members.<br>Leo B. Rasmussen from Jutland became chairman and Arne Crown Rasmussen from Funen became secretary. Members at large were Ebbe Christensen from DELTA Elec-tronics Test, Ole Skjoldstrand from the Society of Electronic Factories and me.<br>The starting point was that ETSI was standardising some types of analogue leased lines under the ONP-LL Directive (Open Network Provision – Leased Lines) which would replace some of the types provided up to now by Tele Danmark, and which we had to provide (they belonged to the minimum set of leased lines along with digital 64 kbit/s and 2 Mbit/s lines which had to be offered throughout Europe). As to these types the new technical regulations for the terminals should just refer to the CTR’s to be issued by the EC Commission based on the TBR’s from ETSI (Common Technical Regulations, resp. Technical Basis for Regulation). I had been a part of the planning within ETSI to prepare these regulations as a member of PT9V in the autumn of 1991.<br>There were four types of analogue leased lines for which this applied viz. 2-wire and 4-wire lines of normal and special quality. There were, however, only two TBR’s, resp. TBR15 for 2-wire and TBR17 for 4-wire terminals. The difference in quality came in what the network operators should deliver (e.g. less variation of delay with frequency for lines of special quality).<br>But in addition to these types Tele Danmark also offered lines with signal conversion, with local quality and with broadcast quality. For these types of lines we had to go in detail with the requirements to the connection of terminals in the new technical regula-tions. At the same time we should adhere to the Terminal Directive from the EC, which said that it was only essential requirements which should be complied with in order to get a type approval for a terminal. Considering leased lines this meant that it was only possible to require such properties that the telecommunication networks were not harmed.<br>This caused many discussions in the group. My starting point was that there should be as wide limits for the terminals as possible, while Leo rather wanted more limitations. As to lines of local quality which were supplied as wire pairs between the customer ad-dresses (and were therefore normally limited to have both end points within the same exchange area) Tele Danmark specified that they were for transmission up to 15 kHz. In real life wire pairs can be used at much higher frequencies, e.g for ISDN2 connections using power up to about 50 kHz (and think of the DSL lines during the 2000es). My viewpoint was that customers should be allowed to use the lines at the highest frequen-cies they wanted, e.g. to an ISDN2 terminal under a PABX but at another address. It should only be required that they should at each frequency send at such a level that cross talk could not disturb neighbouring circuits. It should not be required to select wire pairs for this type of lines.<br>In the IEEE Communications Magazine there had been an article (Chen and Waring, May 1994) stating that the crosstalk attenuation decreased by 15 dB per decade, i.e. with 15 dB from 10 to 100 kHz and again with 15 dB from 100 kHz to 1 MHz. The labora-tory of Jutland Telephones could confirm that this was also the case in Denmark (it had been measured when ISDN was introduced). As a consequence this value was stated in the technical regulation as valid for the range from 15 kHz to 1 MHz.<br>In the spring of 1995 we could send our draft to the TSB and during the summer it was published for comments. The TSB had divided it in 3: TB95069 referring to TBR15, TB95070 referring to TBR17 and TB95071 for terminals on the special Danish lines. Comments were received from Tele Danmark to the latter and then nothing happened for a long time.<br>The reason was internal changes in the TSB. Ernst Hansen retired and Helene Jensen replaced him. She came from outside the TSB and had to work herself in first.<br>In the meanwhile other things happened. The EC Commission had decided that the only case in which the essential requirement for cooperation between terminals via the public network should be applied in connection with analogue terminals was for speech trans-mission between normal telephone sets directly connected to the NTP’s (Network Ter-minating Points). As a consequence the requirements to transmission and to the inter-face towards the extension, such that transmission from a normal telephone was possi-ble, in the technical regulations for PABXes were to be discarded. We had in TB95071 referred to exactly these requirements when writing the rules for terminals for lines with signal conversion from and to the usual telephone signals (ringing voltage and loop in-terruptions) on 2-wire lines.<br>Thus the group was again activated in the autumn of 1996, but in the meantime the two others from Tele Danmark had moved to other jobs within the company and could not continue. I had then to continue as both chairman and secretary. The work to include the requirements directly in TB95071 was, however, rather easy. We had only one meeting and managed for the rest with a few faxes before we could in the middle of November send our draft to the TSB.<br>This did not end our job. It would only be finished when both a Danish and an English version of the technical regulation had been issued. We took this up with the TSB, which were also very responsive to our proposal that the final editing of TB95071 should follow our proposals closer than was done in the issue published for comment.<br>It is lightly ironical that the requirements added directly to the technical regulations were for a disappearing circuit type. Tele Danmark could not any more buy the signal converters necessary between the physical line and the transmission equipment. At the same time the transmission equipment always bundled the circuits in groups of 30 cir-cuits (PCM systems) which had to follow each other. This was not economically sound if you only needed one circuit. The signal converters could be replaced by interfaces to the LLSN (Leased Lines Services Network) which was being introduced and which had the required 2-wire interfaces. But between such interfaces there was still the require-ment that 30 circuits had to be provided in parallel.<br>Tele Danmark had other cards in its sleeve. Digital public exchanges were introduced in great numbers and their facilities could be used. Especially the Centrex offering, a coun-try-wide PABX functionality with no other equipment at the customer address than normal telephone sets, was useful. Using this offering as the main PABX and with cus-tomers having PABXes and normal sets under it the same functions could be provided as with leased lines with signal conversion. In addition the customer should not – as with leased lines – pay whether he used the line or not (leaving it also to Tele Danmark to somehow cover the cost of 29 maybe unused circuits). Now the connections were in the switched network and existed only when needed. When a connection was not in ac-tual use the different parts of it were available for other users.<br>The Part 8 group<br>Part 8 of the Danish requirements to PABXes treated their access to “customer links” or 2 Mbit/s PCM connections to the public exchanges. These are structured PCM systems, meaning that the bit stream is divided in 32 timeslots of 8 bit each, repeated 8000 times per second. Timeslot 0 always has a certain pattern, making it possible for the receiver to recognise it and find all timeslots (and bit). Timeslot 1 to 15 and 17 to 31 are the 30 speech channels in which speech (or tone signals) are coded according to the A-law (in Europe). Timeslot 16 is the signalling channel, in this case using CAS or Channel-Associated Signalling, meaning that every 16th timeslot 16 has a certain recognisable pattern, making it possible to associate the 15 remaining timeslots 16 with 2 speech channels each. Thus each speech channel is accompanied by 4 signalling bit, repeated every 2 ms.<br>The signalling bit are used for line signalling (seizure, release, answer), while the regis-ter signalling (forward signals like dialled digits, type of caller, backwards signals to call for next digit or caller type or to report if the called party is free or busy) were MFC (Multi Frequency Code) signals in the speech channels. Already in 1974 I had worked with this, although on analogue lines, for the DDI (Direct Dialling In) function in EBX 8000 for CTBD and later for the state network of PABXes. From the beginning this sig-nalling was very flexible, as proved when it was discussed with Siemens in 1992. Unlike other PABX suppliers they wanted to receive the signal for type of caller before the dialled digits, but this was no problem, if only they asked for it with a backward signal and then asked for the relevant digits afterwards.<br>In 1990 the PABX-group had issued Part 8 in a “provisional draft” form. It could not remain so. The reason was that there did not exist at that time a real possibility for test-ing whether the terminals complied with the requirements. In 1994 the standardisation of ISDN was so far that it was at least possible to make proper test specifications for the physical connection.<br>Thus, the TSB created in 1994 a group besides the PABX-group to make a revision of Part 8. It did of course mostly consist of people from the PABX-group. Tage Fox Maule from TDBD, Tele Danmark Business Division, became secretary and should edit the draft.<br>While the group was active changes occurred in what should be required from PABXes. It was laid down that requirements should only apply to the connection to the line to the public exchange and be limited to the basic functions. Transmission requirements should be scrapped, as the essential requirement for cooperation between terminals in relevant cases, article 4(g) in the Terminal Directive, should only apply to normal tele-phone sets for live speech when the sets were connected directly to the analogue Net-work Terminating Points. The TSB was also becoming stricter as to what a Technical Regulation might contain. It should in the future be limited to the requirements and the test arrangements making it possible to verify the requirements.<br>The total set of PABX requirements was therefore cut down considerably. There was still a Main Regulation (the old part 1 and 2 with abbreviations etc.) but part 3 about analogue connections to the switched telephone network was replaced by the rules in NET4, edited by the N4G-group. Part 4 about analogue connections to leased lines ex-isted still in 1997 (when I retired) but would then be replaced by the rules edited by the TAFK-group. Parts 5 and 6 about supplementary equipment and functions on analogue connections to the switched telephone network were scrapped. Part 7 about the connec-tion of analogue telephone sets on the extension side was also scrapped. Part 8 about connection to customer links should remain and be edited by this group. Part 9 about connection to digital leased lines was replaced by a reference to the Technical Regula-tions for terminals on such lines. Parts 10 and 12 about transmission in resp. analogue and digital PABXes were scrapped. Part 11 did never exist.<br>As said: Part 8 should remain. And considering that there was no Technical Specifica-tion from Tele Danmark for its delivery of customer links we did for a long time neglect the new guide lines from the TSB and kept an informative chapter with such a specifica-tion. We did of course also advise that Tele Danmark should issue such a product speci-fication.<br>For the rest the job was a careful examination of the old draft and the solving of all doubtful points. A funny aspect was the scope of requirements. No test laboratories were interested in making a set-up for automatic testing of terminals for customer links, the market for tests would be too small for it to pay off.<br>The outcome of this was that when we handed over our result to the TSB in October 1995 there was an informative chapter and the requirements were limited to cover only the physical access, as it could be tested in the set-up already used for test of ISDN30 (or primary rate). Whether the signalling worked towards the public network we left to be based on suppliers’ declarations.<br>The TSB did not like this – they had no authority to implement such a rule. They did neither like the informative chapter.<br>In 1996 the old telephone administrations had been reorganised into functional entities in the Tele Danmark concern and the work with specifications was arranged in a new way. One point was that John Hartrup in NIS (Network Infrastructure Switching) should make a technical Specification for customer links. He later moved to the regional or-ganisation for Copenhagen and the work was taken up by Jürgen Endres in NUS (N Development S) with assistance from Harly Østergaard Jensen, U (Development). They were both members of the new PABX-group under the TSB. This resulted in TDK-TS 900234 of May 1997.<br>This meant that the informative chapter in the Technical Regulation could be replaced by a reference to this TDK-TS. Thus, in June 1997 the TSB could finally issue TB97087 with access requirements for terminals on customer links. I did not see it before my re-tirement, so I do not know the result of the considerations concerning the scope of test-ing.<br>The BELA-group<br>A group was formed within Copenhagen Telephones in 1995 to describe limitations for electronic solutions in the access network. The group should compare such solutions with having a copper pair to the customer.<br>Copper wires can do much more than transmit the necessary frequencies for speech in the range 300 to 3400 Hz. Their attenuation does not increase steeply above this band, in stead it increases gradually with the square root of the frequency. If the attenuation is 3 dB at 1000 Hz, it will be 6 dB at 4000 Hz etc. Therefore one can work with rather high frequencies such as ISDN2 connections requiring transmission up to about 50 kHz or even ADSL (Asymmetrical Digital Subscriber Line), using the copper pair to trans-mit TV-signals at several MHz to the customer.<br>An electronic solution might be used in a place where there were no more copper pairs to avoid laying a new cable, by e.g. installing electronics at the two ends of the fully used cable to double its capacity by transmitting ISDN2 signals on its pairs and translate these signals to normal copper wire signals at each end of the “narrow” cable. But this will limit the transmission for the connections to the speech band.<br>This is not the only limitation. The copper wire is also used to feed the telephone instal-lation such that it will work even if there is no electrical power in the house. It might still be possible to feed two installations via a single pair, but if more connections shall use the same pair a local power supply is necessary including a battery if the installation shall be independent of the electrical power system. However, when there is power via the copper pair one can also feed with one or the other polarity and this can provide ad-ditional signals to the installation, e.g. a polarity reversal when the called party answers. These signals or short interruptions of the feed current are impossible to use with most electronic solutions.<br>An important point is that customers subscribing to one or another signalling possibility shall not be transferred to electronics which can not transmit the signal. In order to avoid this one must at first map which limitations the different electronic solutions will imply. We choose in the group to make a list of the basic services, such as polarity reversal, and whether they could be delivered by the different electronic solutions. We then made a list of customer services and their need for the basic services. The sales staff needed such a list when they discussed with the customers.<br>The work led to a report, the last one before Copenhagen Telephones was disbanded at the end of 1995. Within the Network Division Johnny Damgaard continued the work and added in 1997 another pair of solutions and their limitations to the report.<br>The Housing-group<br>In the middle of 1996 all delivery of telecommunication services in Denmark was liber-ated. The laws were not quite ready, this happened only a year later, but it was obvious that there would be several requirements to Tele Danmark as it was the dominating sup-plier. One was the requirement that others could demand to get room in the buildings of Tele Danmark for their equipment to obtain “interconnect”, viz. the coupling between the networks.<br>It would be a matter of consequence that there should be rules for such interconnections. The other operator of mobile telephony, Sonofon, already had this type of interconnec-tions, otherwise their subscribers could only connect to each other, not with the users of the fixed network. In that case there would be only a few people preferring Sonofon, even if calls in their network were free!<br>The starting point might have been similar interconnects as those Tele Danmark had with telecommunication operators abroad, i.e. connections via transmission systems between equipment at the premises of each operator and with agreed protocols for transmission and signalling on the connection. But as stated the requirement here went further, the operators should have a right to install their equipment in the buildings of Tele Danmark.<br>Different groups were formed to consider the consequences for Tele Danmark and pro-pose rules for “interconnect”. I participated in a group evaluating the consequences of the requirement for collocation and we met repeatedly during 1996.<br>Our starting point was that interconnect should take place at a rather high level in the network, just like the connections to operators abroad. A number of exchanges through-out Denmark were chosen in which other operators could have their equipment installed and be connected to the Tele Danmark network.<br>A problem was and is that Tele Danmark has obligations towards the Danish state con-cerning the safeguarding of telecommunications under all circumstances. For this reason it can not be allowed for everybody to enter the buildings. This caused another require-ment to the buildings where interconnect should be possible: They had to be divisible in an area with access for other operators and an area exclusively for the Tele Danmark staff. Collocation was therefore divided in three classes, either each operator had his own locked room with a common access from the outside, or the operators had a com-mon room with access from the outside, or – as the third possibility – the equipment was placed in the same room as the equipment of Tele Danmark. In that case it was required that there were always personnel from Tele Danmark present during access.<br>Luckily there were already examples of other operators of especially data services hav-ing their equipment collocated with Tele Danmark equipment. The agreements about this was a good starting point.<br>However, collocation could not be limited to those few places in Denmark. A new op-erator might be a supplier of TV distribution who extended his plant to replace the Tele Danmark access network, i.e. the network between exchange and the single user. Such an operator needed access at a subscriber exchange and the rules had to adapt to this. In fact it applied not only to the need for collocation (where access to a transmission sys-tem in most cases could cover the needs) but also the subscription price. Tele Danmark had to realise that the large expenses in the access network disappeared in this case, such that the new operator could charge the fixed pay for the subscription. But there was compensation: The new operator was responsible for the whole access network and Tele Danmark should not any more make up bills to all subscribers, the new operator was their sole representative.<br>In the end of 1996 the work was halted as there was doubt about the legislation and its consequences for Tele Danmark. Early in 1997 the chairman of the group was delegated to Newtelco in Switzerland, which had been partly bought by Tele Danmark, and he was later employed by Newtelco. Therefore the group was inactive up to the time I retired.<br>The MCID-group<br>I became part of another group in 1996 and 1997, but not as a proper group member. The group should make proposals for a change of the alarm service such that it would not be necessary to have a specially developed program for the AXE exchanges from LME to run it. MCID means “Malicious Call IDentification”.<br>The reason for not being a proper member was that this matter applied to exchanges and my department, NUI, had nothing to do with them, they were the field of NUS. But con-sidering my past I knew about exchanges and signalling, and therefore the group wanted my active participation.<br>The specially developed program saw to it that when there was an alarm call the AXE exchange switched in a tone receiver in parallel on the line during the whole call. When the alarm centre pushed a code on their telephone set the receiver caused a signal to the supervision centre of Tele Danmark, where all exchanges are supervised at all times, printing out the number of the caller. The staff here could find the address of the caller and – when the alarm centre called – tell it.<br>The standard program in AXE for MCID did not switch in a tone receiver constantly. It was only called in (and sent a dial tone) when the called party (i.c. the alarm centre) sent an R-key signal, a short interruption of the current in the subscriber line. The develop-ment of the special program for Denmark had a higher and higher cost for each new program package for AXE, so it would be an economical advantage to use the standard program.<br>But this was not only about an extra push of a button in the alarm centre, a push already introduced in several places in Denmark. Many alarm centres used a PABX, possibly connected via a 2 Mbit/s link to the public exchange. Thus, the real problem was whether the used PABXes could send the R-key signal to the public exchange? An R-key signal was already used in PABXes (from long ago) to call for a new dial tone to start a call-back and transfer. The PABX would thus have to, after having received an R-key signal, perceive a special dialled code as a message meaning that it should send an R-key signal to the public exchange. There would then come a dial tone from this exchange and the code for printing the calling number could be dialled. The user at the alarm centre should in this way both push the R-key and dial the special code in addition to what he was used to. This was not very ingenious!<br>In addition to this all alarm calls in Copenhagen go to the fire department, who transfers the call to the police if it is about other than fire or ambulance services. In this way the police might have to send three codes after the R-key signal before the last one gave a print-out of the calling number!<br>One of the solutions considered was to make all connections to alarm centres into ISDN connections. In this case there would at all times be a special signalling channel avail-able independent of the speech channel to send the necessary signal from a function key during the whole call. However, such a change was not cheap and might require new PABXes in the alarm centres.<br>Finally, the solution could not just be a proposal from the group introduced over night. It had to be field-tested before coming in actual use. This had to wait for the next pro-gram package for the AXE exchanges to test in connection with the standard program. Therefore the group had not finished its job when I – along with several others – left Tele Danmark.<br>Retirement<br>In the summer of 1997 the message came from the CEO of Tele Danmark: All employ-ees born in 1940 or earlier could terminate their employment at the same conditions as if they had been fired. This suited me fine, I would shortly be 65 years of age, the normal pension time if I had stayed with Philips, and the economy was OK. Further, like all oldies I could expect that if I did not resign, I would be fired! Thus, by the end of Au-gust 1997 my professional career ended...

Revision as of 13:47, 7 December 2009

Contents

Preface


The title is somewhat misleading as I left Philips in 1991 and also because I had not much to do with the inner workings of the exchanges since the middle of the 1980es. Thus, this account of my finishing years working with telephone exchanges is rather an account of the many groups in which I was involved during these years.

The PABX group


Since 1988 I was a member of the Danish PABX group, which was formed by the telephone companies of that time (covering Copenhagen, Funen, Jutland and Southern Jutland – the last one a totally state-owned entity as a remnant of Southern Jutland being under Germany from 1864 to 1920). The PABX group had as its task to prepare for the liberalisation of the PABX market.
Small terminals had been liberalised for quite a while and Philips had obtained approval of e.g. Sphericall and Multicall according to these rules. Later also telephone sets had been liberalised. PABXes would come next.
At the same time the approval process was removed from the telephone companies, a policy later followed throughout Europe. It was inappropriate that one of the players at the liberalised market should be the approval body for the equipment their competitors would sell. Thus approval was transferred to a body directly under the relevant minister, The Telecommunication Control Board.
But measurements to control that terminals performed as required continued to be at the laboratories of the telecommunication companies. The former State Telephone, Copenhagen Telephone and Jutland Telephone had approved laboratories for these tests. Later also laboratories outside the telecommunication companies were added to this list, e.g. DELTA Electronic Testing. Only test reports from the approved laboratories were accepted by the approval office.
The PABX group was formed to prepare the liberalisation of the PABX market. Right from the start it had representatives from the suppliers, i.e. from IBM and LME. IBM soon left the group and from its 12th meeting there were representatives from Alcatel (or then: Kirk’s Telephone Factory) and Philips. The structure in the technical requirements was already rather clear: There was to be a part 1 with a survey of all parts, a part 2 with definitions, and parts 3 to 12 for the different interfaces to the public exchange and the extension side and for the transmission through the PABX.
The first job was of course to get all these parts ready in time. It really stressed the group when the time for liberalisation was set to July 1990, for then all parts had to be ready. We lived up to this.
During the following years, with about one meeting per month, we adjusted the single parts according to the experience with the use of the requirements and according to the development at other points, as e.g. the decisions by the European Community as to what it was allowed to require. I remained a member although I changed company during these years from Philips to Copenhagen Telephone Business Division (CTBD) to Copenhagen Telephone Tele Division and finally to Tele Denmark Networks.
An example is part 3, “Connection Requirements for PABX Interfaces to Analogue Lines to the Public Switched Telephone Network”. When we had to revise it, i.a. due to the Terminal Directive from the Common Market describing at which areas requirements were essential and thus admissible, I got the job as editor. I made an effort to write a good chapter with technical information, as a description of the properties of the analogue subscriber network in more popular terms than in the old Circular 2 was missing. But the Control Board wanted only the requirements to be included in part 3, the information belonged to a specification from Tele Denmark of its deliveries to its customers. A very justified point of view, except that Tele Denmark was not yet so far that it could produce declarations of its offerings to the extent the market would require when not only the terminals but the whole network might be supplied by others.
Tele Denmark was busy writing declarations (see later under the PSTN group) and the information chapter of part 3 was offered here as an appendix to the declaration. However, when the declaration was issued as a Tele Denmark specification (TDK-TS 900221) it was not included. I do not know if this chapter was ever used at all!
It went even worse with the rest of the revised part 3. The N4G group was set down at the same time to revise the Danish input to the ETSI (European Telecommunications Standardisation Institute) specification ETS 300 001, such that it also covered require-ments to PABXes (see later under the N4G group). With only half a year between our revision of part 3 and the deadline for the work in the N4G group the Control Board decided that the revised part 3 should not be issued! A pity for us, but the discussions in the PABX group were a good background for discussions in the other group.
The PABX group had its 100th meeting in the end of May 1995 and the chairman, Jørgen Hess Kristensen, and the secretary, Harly Østergaard Jensen, throughout its whole time were duly celebrated. Driving home from this meeting I heard on the car radio that Prince Joachim was engaged to Alexandra from Hong Kong.
The group held a total of 112 meetings. It was in general the Control Board who formed the groups writing technical requirements, and the PABX group originally formed by the telecommunication companies was only surviving as a historic remnant. But now this came to an end. The Control Board formed a new group and Tele Denmark stopped the old one. The new group kept the same chairman and secretary and they also wanted me to join in. I was at that time in the Tele Denmark Network Division in a subdivision for network and service development, platforms for the infrastructure (NUI), and my boss decided that PABXes were outside our area, they fell under exchanges. As a consequence the 3rd representative from Tele Denmark in the new group became Jürgen Endres from NUS, switching platforms, with whom I had worked together in the N4G group.

Murphy’s Law at ISS 90


In 1990 the International Switching Symposium was held in Stockholm and I applied as so often before for permission to participate. At first I got the green light, but as the trip came nearer and the money for it should be granted, Afzelius (who had replaced Øberg as head of Philips Tele Nordic when Øberg went to CTBD) got cold feet. In the end he gave in (I would have been a participant in any case even if it had cost me a vacation week) and in the last days of May my wife Jytte and I went to Stockholm for a week with a lot of technicalities but also a week with a lot of nice events besides the symposium itself.
The suppliers did what they could to give the delegates some nice evenings. Monday there was a reception at the Town Hall, in tents erected in the gardens facing the Mälaren lake (sponsored by Philips), Tuesday we went on board the small steamers from Gamla Stan to the fortress Waxholm in the archipelago, having dinner on board and coffee in the fortress yard while a choir sang and there were fireworks (sponsored by Siemens), Wednesday we were in the Globe for a show and a Viking-dinner, with the cooks working on platforms just under the ceiling, 100 m up, from where they were hoisted down when dinner should be served. Later there were indoors fireworks and a performance by Lill Lindfors (sponsored by Televerket and L. M. Ericsson), and finally Thursday Alcatel ITT had booked the whole Skansen, open-air museum and zoo, with dinner and coffee in the various old buildings – and again fireworks, this time over the harbour, a fine sight from Skansen!
From the symposium proper only a little glimpse of Murphy’s Law (if anything can go wrong, it will). One of the assemblies had Sterndorff from the Danish Telecom as moderator. During the first speech everything went wrong, the dias were in the wrong sequence, were upside down etc., so after a few minutes Sterndorff interrupted the speech to let the speaker have time to arrange his dias better. And Sterndorff announced the next speaker, mr. Murphy. The audience roared with laughter while Sterndorff looked like he did not know why!

Autumn 1990, ETSI PT26


Back in the 1980ies Jørgen Michaelsen, with whom I had worked on EBX and TBX since 1974, had left CTBD a little later than Oksholm to become the first head of the cable-TV group. In the spring of 1989 he switched offices with Bent Nielskov to become head of CTBD, only to be replaced after a very short time by Niels Øberg from Philips, my boss there through 10 years. Michaelsen choose to leave CT completely to go to Fischer and Lorenz Telecommunication Consultants as head of their subsidiary European Communication Consultants.
After a meeting in the PABX-group in Aarhus in June 1990 I met Michaelsen on my way back to Copenhagen by plane. He said they had now and then tasks in F&L where my knowledge might be useful. Would I be interested? Certainly, but I would have to discuss it at Philips, both if it was possible at all and all details if it came to something.
It was possible and the first task was to represent F&L in the ETSI (European Tele-communications Standards Institute) project team 26. The job turned out to be the creation of a data bank listing comments to access requirements for the switched telephone network throughout Europe and from all European countries, and with possibility for sorting according to different criteria. Luckily I had tried this sort of work on a PC in my office, and this helped us a lot. I wrote the following article in “IT-Nyt” about it in 1991:
HARMONISATION OF ACCESS TO THE TELEPHONE NETWORK
The most widespread possibility for connections from anybody to anybody is presented by the public telephone network, the largest machine in the world with more than 800 million access points around the globe.
It is no wonder that the EC (European Community) would like to have the access to the network harmonised, i.e. governed by the same rules and technical conditions in all Community countries. It is also no wonder that this work will take time, as the reason for the extent of the network is that it is old and has therefore grown in a little different direction in each country. A change in a common direction should not mean a decline in the quality we are used to: A connection when we want it (if the called party is not busy himself) and a speech quality as if the other party is in the next house even if we know he is halfway round the globe.
But – is the network not harmonised already when one can phone from country to coun-try without any problems?
No. There are international agreements but they do not govern details within the single countries. An example: There is an international agreement on the level of the electrical signal delivered at the border for a certain sound pressure at a microphone on the network. And there is an international agreement on the sound pressure supplied by a telephone on the network for a certain level of the electrical signal received at the border. The aim is of course that the speech can be perceived no matter how bad the micro-phone and telephone are placed in the respective networks. Long lines with a large attenuation must not prevent a reasonable quality. But it was a national matter if a large amplification in the microphone was chosen with a corresponding large attenuation in the lines, or if a microphone with a smaller amplification (and maybe a better sound quality) and a smaller attenuation in the lines was chosen. This implies that telephone sets are not harmonised. A set from the latter country will not work satisfactorily everywhere in the first country and a set from this country will be too powerful when used in the second country. This is not just a question about the use of different plugs in different countries, for this was easily handled by a variety of adaptors!
ETSI has consequently continued a work that was already started to get at least the technical conditions for access to the telephone network in the European countries described in a uniform way, as a prerequisite for mutual approval of terminals for this network. The aim is a technical harmonisation such that an approval of a terminal in one country makes it approved throughout the Community. The first step should be that if you present the approval body in Denmark for a measuring report from a Greek laboratory (approved by the local authorities to do such measurements), which shows that the equipment complies with the Danish regulations, the Danish approval body approves the equipment for use in Denmark.
In fact the Danish approval body is already today so liberal that foreign measurement reports are accepted (however maybe not in Greek!) from approved laboratories as a basis for type approval in Denmark. At this point certain Community countries are very bureaucratic and maintain with a long handling time at the authorities still a “technical trade barrier”. Possibly the Community should put these countries in their place by introducing the same rule as within the power network: Equipment may be sold as soon as the supplier has notified the authorities of its existence. The authorities can afterwards buy samples and test them and if they do not comply with the requirements the equip-ment must be withdrawn from the market, including already sold units.
As to Denmark we had from the early 1970ies, when the first terminal types for use under the telephone network could be approved for sale on the open market, the rules in Circular no. 27A from the approval board. Today (1991) they are issued as TI-publication 11/89 and this is the Danish input to the work at ETSI.
TI-publication 11/89 comprises only access requirements common to all terminals on the telephone network. Referring to the 7-layer model of ISO (International Standards Organisation) only the 3 lower layers are treated: the physical, the link and the network layers. In other words the items necessary for the connection to the network and for signalling to it to get the wanted connection, resp. to receive a ringing signal from it and answer it when another has called. There is no treatment of the higher layers, i.e. items necessary for the network services. Thus there is no treatment of the conversions mentioned above between sound pressures and electrical signals (this is treated in the technical requirements to telephone sets, TI-publication 4/89) and likewise no treatment of e.g. which signals a modem for a given data speed shall use.
The work towards a common description was taken up before ETSI was established, and this is probably the reason why the preparations for ETS 300 001, as it is called and where ETS means “European Telecommunication Standard”, continued at Deutsche Bundespost in Darmstadt. Here each item is taken up separately, e.g. dialling with dial pulses, and put into one chapter with values for all countries.
The proposed standard, prETS 300 001, was sent into a public hearing in two steps, chapter 1 to 4 in January 1990 and chapter 5 to 10 in April 1990. Considering that it was not just a proposal for a standard which one could follow or not, but for a common European binding normative rule under the name NET 4, where NET means “Norme Européenne de Telecommunication”, it was also sent in hearing outside Europe. Comments should be at ETSI early October 1990. ETSI can approve an ETS, and if it has been stated before that it may become a NET, a Community Body can make it applicable in all Community countries. In Denmark we discriminate in this way between standards and norms, while the Community calls them type 1 (our norms) or type 2 (our standards).
The Community is by the way leaving the designation NET, replacing it with CTR for “Common Technical Regulations”.
ETSI expected many comments to the proposal, which has about 1100 pages. There were indeed almost 4000, of which many were completions of values in schematics and the like. Therefore ETSI had prepared a project team, PT26, to sort out the remarks and prepare answers to them. However, the national values should be respected, in this first turn each country shall control its own values, whether it is values in a schematic or a remark.
What differences are there in the national values?
They consist both of items for which some countries have rules while others do not, and of items with differences of more or less significance. Each value is of course important for the country stating it, but it can be difficult to understand for others. And of course a country can find it a nice technical trade barrier to have a deviating value here and there.
Some countries have found it necessary to control the DC current through a terminal in such a way that if more terminals are used in parallel the current from the telephone line is equally distributed among them. Denmark has no such rule. If necessary it shall be a rule in the requirements to such terminals which are likely to be used in parallel and which only get their power from the line. Thus, in Denmark such a requirement is not found in TI-publication 11/89 or in prETS 300 001, but only in the rules for telephone sets, TI-publication 4/89, as sets are now and then used in parallel.
An item where all countries have values is the input impedance when a terminal is in use. Most countries have 600 Ohm as their normal value, but Germany and Switzerland have a different impedance and have stated such tolerances that a terminal with 600 Ohm impedance will not comply!
All requirements are accompanied with a description of how they are verified. This is a general rule in ETSI standards: If it can not be measured there shall be no requirement. Many measurement descriptions in the Danish remarks are copied from Circular 27A. They could as well refer to figures in the common text.
A terminal supplier wanting to sell in more Community countries must either produce several models or design one model with flexibility so it can be approved in more countries. ETS 300 001 does not change this. Maybe it will be possible later to weed out all non-essential items and simplify the rest.
PT26 should as said not discuss the national values but sort them out. PT26 consisted of 4 persons, 3 technicians from the telecommunication administrations in France and Great Britain and from F&L, while the 4th was the editor who had at Bundespost shaped the proposal. In 9 weeks from the middle of October 1990 we assembled the comments to the national values and sent them to the countries (a comment from France to a Danish value should only be part of the final issue if it could be accepted in Denmark). Then we treated all the comments to the general text and made a proposal to the ETSI body responsible for the standard whether they should be accepted or not (and especially in the latter case: why not). We were ready within the time frame scheduled for us.
What has happened since then?
The ETSI body found at its meeting in March 1991 that many general comments were such that another hearing might be required before acceptance. Therefore the standard will first be issued with only correction of obvious typographical errors and the like. Other items have to wait for a new issue.
It is to be expected that this issue will also limit the national freedom of expression. On many items each country has used its own “Circular 27A”, but the result would not change if a common text and figure was used. The Danish measurement methods have been mentioned.
Four new project teams have been proposed to evaluate how to go on. One shall consider the further harmonisation of ETS 300 001. Another shall consider attachments different from those in ETS 300 001 (some countries have other signals between a PABX and the public exchange than between a telephone set and the public exchange, and even if you do not have this you do not connect a PABX via a row of plugs. For a PABX there is also an issue of overhearing between terminals). A third team shall consider which requirements in ETS 300 001 are essential in the Community sense (the others must go) and a fourth team shall supervise and coordinate the other three. In total it is expected that about 30 man-months will be spent just on these preparations within the next half year.
But let us now see the first issue of ETS 300 001, the combined catalogue of the technical requirements within Europe for approval of equipment for use under the telephone network.
What does all this mean for the users?
Not much in the first instance, as the rules are not directed at the users. They shall only be sure that the equipment they buy may be connected to the network by checking that it is approved.
ETS 300 001 and the mutual approval of test results is of more significance for suppliers, as they can easier adapt to the rules of various countries and have their terminals tested at the nearest laboratory, maybe even according to the rules of more countries at the same time. This might result in cheaper terminals.
---
F&L had rented an apartment in Vence, 35 km from Sophia Antipolis, the home of ETSI, for an employee working down there. It was lucky for me that I could use it during my stay in the autumn of 1990. It was a beautiful trip each morning and evening to and from work, via small mountain roads through woods and small towns. From the apartment you could see the coast 10 km away, so while I had my breakfast I could see the sun rise over the Mediterranean. To the left one could see the mountains behind Nice, to the right the holiday apartment buildings near the beach between Nice and Antibes.
The weather was not fine all the time. Shortly after my arrival there was rain in torrents one afternoon. It rained also when I went home, and in one place the water ran from the mountain to the left across the road in a depth of maybe 30 cm towards a house on the other side of the road. There the water ran to both sides through the garden of the house to a river. I sneaked along but further on my way the road was blocked, the river had flown over on its way to Cagnes-sur-mer. Well, I could turn left via a higher road past St. Paul and home.
I had my family (wife and daughter) on holiday during a fortnight. They would not fly the direct route from Copenhagen to Nice, but took the train. However, there were delays, so they did not reach the right train from Bale in Switzerland. My wife had the telephone number to my office, so she called well prepared to talk in French when the phone was picked up. She was very relieved when it turned out that the number gave direct access to my office and I answered the call. Well, I got the new arrival time in Nice and could fetch them towards midnight. We had a nice holiday, also because I had collected the days off granted by ETSI for this fortnight. During weekends I had already been around, so I knew where to find the nice places along the coast and inland.

From Philips to CTBD


After Øberg had left Philips in June 1989 to become head of CTBD there were still technical questions to be discussed before SOPHO-S could be introduced at the Swedish market. The sales- and service-organisations were ready in 1989, the approval of SOPHO-K was obtained and the sales had started, but the main piece was still missing. This was – as said in an earlier essay – brought in order in the autumn of 1989 and then the Swedes could by and large run the show themselves. They had to, anyway. During 1990 they were often visiting us with customers to see a working exchange and listen to customer experiences (at Philips themselves), but that was all.
Tele Nordic got a new head after Øberg. This was Afzelius, who after a period at Philips had been a manager at Adidas, then came back as head of the service organisation at Philips and now became head of Tele Nordic. It was my impression that he never understood much about telecommunications, except for mobile telephony where our factory made the handsets which were serviced by his organisation. The sales group for mobile telephones was joined to the sales group for telecommunications, for which Finn H. Nielsen continued as head, and Afzelius was head of both sales groups and Tele Nordic.
When we heard from Sweden that the first orders for SOPHO-S exchanges were received I found it time for Philips Sweden to pay part of the cost of Tele Nordic. It could not any longer be a problem for Denmark alone as during the start phase (where Denmark was the only Nordic country with some sales of telecommunication equipment). It would also show whether they would not just say nice words about my assistance but literally speaking found it valuable. Thus I proposed to Afzelius to write to the Swedes that if they wanted me to come in the future they would have to pay for it. I never saw an answer to this and I did not participate in meetings in the Swedish standardisation group any more.
So generally spoken my work thinned out during 1990. I heard a rumour that Afzelius should have asked others in the group “what does Poulsen really do?”. Therefore it suited me fine to work for Fischer and Lorenz in PT26 during the autumn.
Within Philips International Jan Timmer had taken over as president (CEO) after van der Lugt, who was only interested in the football club PSV. This was probably not the only reason why the company had some years with heavy losses, but he was as said replaced and Timmer started a tough program to thin out the company and make costs fit better to the sales. In Holland everybody above 55 years of age had to go, except Timmer, he was 56! Several good people in Hilversum were forced out at this time. I was myself 58 years of age now.
The program had of course also consequences in Denmark, and our secretary Betty Bjerre had repeatedly during the autumn warned me that I should not feel too sure about my job in spite of the more than 30 years I had been with the company. I was of course aware of this but carried on as if nothing was wrong.
But then one day in January 1991 the new second in command in the mobile telephone group came to my office to tell me that they did not need my work any more. They had contacted both Øberg in CTBD and Michaelsen in F&L, who would both like to talk with me about employment. I immediately told Betty about it and phoned also to my home, for we had of course discussed the possibility which had now turned into reality.
Well, there was nothing to be done other than talking with Øberg and Michaelsen, and this I did over the next couple of days. In CTBD I could be employed full time at the same gross salary as at Philips (luckily I had used all possibilities for extra contributions to my pension) and I knew many of my future colleagues (who, it turned out, all greeted me welcome to the firm as if they had longed for me). Michaelsen could not promise a full time job, but the hourly wage was higher. I was much in favour of both but had to make a decision. I choose CTBD, who should then send me a contract, but this took time. Michaelsen was disappointed that I did not go to F&L, Allan Fischer-Madsen, one of the owners, told me when I saw him at a later occasion.
Philips got somewhat impatient during February. I wondered why I had not received a written note of discharge, so I knew what to expect. I got the reason for this from the HRM or personnel manager, Henning Friis, when we met towards the end of February: They would only give me the note in writing when my employment in another firm was in place! They believed me once more as to why this had not yet occurred and my discharge payment was agreed. I had a right to a payment of 3 months salary when I was fired and got 6. My wife was not quite satisfied after so many years, but I found then and now that any payment above the obligatory was a gift. It should be noted that in addition to the 3 months salary I would have had a right to receive my usual salary during up to 6 months unless I got an employment. Of course Philips did not have to pay this when I went directly from them to CTBD.
Finally the contract from CTBD came early in March. Its content was as agreed and I was to be employed in the group for technique and market, EM, with Erik Andreasen as my boss. In the group I was to be under Mogens Thomsen, who I had known since the days of the TBX exchanges.
Things have to occur in the right sequence, so I found I should be fired from Philips before I signed the contract. The next morning I therefore called Hanne Larsen in the personnel office and I could almost hear her jump in her seat when I asked to be fired, as now the letter from CTBD had come. I got the discharge note, signed it, signed the contract with CTBD and was now definitely on my way out from Philips after 32 years and 2 months.
Finn arranged a farewell reception just before Easter and Johnny Frost had arranged a surprise. They had decorated with an operator’s desk from EBX 8000 and I was asked to use it. When I pressed a key smoke billowed from it!
The group had also collected for a gift and had discussed it with my wife. She had proposed a lawn mower with bread-motor (a hand-driven one), as the one we had was 27 years old. Having only a small garden I was not interested in something electrical or motor-driven, I must go on having some exercise!
At the reception I heard about the reaction of Betty. In spite of her many years in Denmark (after originally coming from England) she was not familiar with all our special expressions so when she heard that a mower with bread-motor was wanted she had asked if such one was not far too expensive? The only place she had heard about bread and motor was when Øberg had made a Volkswagen Van (called a bread in Denmark) into a mobile holiday home. So it was a quite natural question – but we laughed anyhow.
During January or February there had been advertisements for personnel to the new ERO, European Radio Office, which was to start in Copenhagen. Betty was interested and Øberg helped her with contacts to the right persons. She got a job and had everything shaped up on the day of my farewell reception. So she went to our boss to an-nounce that she quit. It made just as much impression as the hole left when you take a hand out of a bucket of water, she said. Philips also gave her a farewell reception in the middle of April, in which I took part, for the first time with the correct CT tie around my neck, a gift from Erik Andreasen.

Autumn 1991, ETSI PT9V


After a few months at CTBD, Michaelsen at F&L had again a job for me. This time it was PT9V at ETSI, which should prepare the standardisation work for leased lines under the ONP-LL directive (Open Network Provision – Leased Lines). The directive aimed at the delivery by all network operators within the European Community (EC) of leased lines (permanent circuits) with certain properties between arbitrary locations within the EC. ETSI was to write the necessary standards for these circuits, divided in the characteristics of the circuit, the interface to the circuit and requirements to the terminals. The latter was divided in two parts: a standard with a broader aim and a requirement specification based on the terminal directive and thus limited to the essential requirements as to user safety, the survival of the network etc. The requirements did not comprise any demands that the terminal should be able to provide any service to the user, caveat emptor!
I wrote an article about this work at ETSI in IT-Standardnyt, as follows:
LEASED LINES WITH A GUARANTEED QUALITY FROM JANUARY 1ST 1993.
From January 1st, 1993, the telecommunication administrations in the EC shall be able to provide leased lines (LL) with a guaranteed quality between arbitrary locations within the EC. Leased lines are called APL in Denmark (for Subscribed Private Lines). The requirement to deliver these lines is part of the harmonisation and liberalisation within the EC of the telecommunication services, also called ONP, Open Network Provision.
As to Denmark the required types of LL correspond to some of the types of LL which can already be provided within the borders of Denmark. The requirement to a guaranteed quality is, however, new. The requirement for a digital 2048 kbit/s LL is new, to-day the users can only get an LL with 1984 kbit/s.
What is ONP?
In 1987 the Commission of the EC issued the Green Book about the future policy for telecommunications. It did not only contain requirements as to harmonisation of the access to the network, with the aim that a terminal (a telephone set, a modem or the like), which can be used in one country, can also be used in others, but required also of the telecommunication providers (who still were monopoly companies) that their networks should be accessible for other service providers and the services delivered by them. This was formalised in 1990 and called ONP. The details were worked out in General Directorate XIII under the Commission, supported by an “ONP-committee”.
Several of the properties which the ONP shall make available for the service providers must wait until the network can supply them. The free use of LL, such that a user e.g. can provide unused capacity to a third party, need not wait. Therefore the Commission has started with a directive about this.
The Directive about ONP-LL and LL types.
The ONP-LL Directive states the conditions for supply and use of these circuits. An important point is that it must be possible to go to one telecommunication administration to get an LL between two arbitrary locations within the EC. This administration shall then arrange for the agreements with other administrations. A probable interpretation of this will be that one shall go to the administration which shall supply one end of the LL.
This article will not deal with supply and use but with technical questions. The Directive addresses technical matters in its requirements as to which types of LL shall be possible between locations in the EC and when, and in its requirements to the properties and interfaces of these LLs.
The Directive requires the following:
LLtype Interface Properties Time scale
Normal analogue 2 and 4 wire M.1040 January 1993
Special analogue 2 and 4 wire M.1020/M.1025 January 1993
64 kbit/s digital G.703 G.800 series January 1993
2048 kbit/s unstructured G.703 G.800 series July 1993
2048 kbit/s structured G.703/G.704 G.800 series July 1993
 

M.1040 etc. are existing CCITT recommendations, i.e. internationally agreed recommendations, and it will be obvious that among their possibilities only those shall be included, which are already in widespread use in Europe.
An important part of the writing of standards is also that every requirement shall be followed by the description of a measuring method to verify it. There are no measuring methods described in the said CCITT recommendations.
National rules
The lease of fixed connections in the telecommunication network is not a new idea and all countries have laid down rules for the terminals which may be connected to the leased lines. In Denmark the rules for connection to analogue lines are found in Publica-tion 12/90 from The Telecommunication Survey Board. This publication contains the old Circular 12A of February 1979.
Rules for the connection to digital leased lines in Denmark have only been issued in draft form, but this has not delayed the use of such lines.
It is a typical trait for all countries that the rules only apply to the terminals. The quality of the connections is hardly treated. The Danish Teleadministrations are, however, an exception as they indicate i.a. the bit fault rate for digital leased lines. Copenhagen Telephones state in their product catalogue that it is less than 1 per million bit for 64 kbit/s and less than 1 per 10 million bit for 2 Mbit/s. They issued also already in 1979 internal directions for the quality of analogue leased lines and have prepared a “Circular 12D” with specifications of the transmission over analogue leased lines in the three available qualities, local, normal and special, of which the two latter correspond to the types stated in the ONP-LL directive.
Quality requirements to Leased Lines.
The standardisation covers three topics. The first is the requirement to the transmission through the leased line.
As to analogue leased lines the directive states that the rules shall refer to the CCITT recommendations M.1020, M.1025 and M.1040. The purpose is that there must not be added requirements or tightened requirements in relation to what is already stated in international standards. It will, however, be necessary to add a requirement to the total attenuation and a few other items.
As to digital leased lines the directive refers to the CCITT recommendations in the G.800 series. They state requirements to the precision of the bit rate, to small variations in it and to the bit fault rate.
The directive requires a continuous control of the quality, but this is impossible to do for analogue leased lines and for the first two types of digital leased lines in the table above. The reason is that the total capacity of the connection is available to the user.
Only the last digital leased line, 2048 kbit/s structured, contains extra bit for a continuous quality control. This type of leased line was not included in the first draft for the directive, it was only included after discussions in the committee within ETSI, which is responsible for the ONP-LL standards.
Details can be found in the “Draft ETR” made by the committee as a basis for the standards.
Interfaces to the Leased Lines.
In addition to the delivery of connections with a stated quality the teleadministrations shall give access to the leased lines via standardised interfaces.
This means that the socket or the like to which a terminal is connected shall be the same throughout the European Community.
More details are also for this aspect found in the “Draft ETR”.
Access requirements to terminals for Leased Lines.
The third aspect addressed by the standards is the requirements to the terminals. They shall be as few as possible, in agreement with the so-called terminal directive of the Community. This directive prescribes that only essential requirements may be included and they apply to:
• The operational safety of the telecom network
• The cooperation with other services
• Data protection (personal data)
• Safety for personnel
• Safety against disturbances to or from other equipment
The cooperation with other services and data protection will not be included in the standards for leased lines. These lines shall deliver a given transmission capacity with a given quality, and a possible supervision shall only ensure this. Whether the contents may be transmitted legally is no concern for the standard.
The two last items are covered by the low voltage and the EMC directives. The standards for leased lines shall therefore not consider these points.
Thus only the requirement as to the operational safety of the telecom network is left, meaning requirements to the maximum signal strength to avoid disturbances of other circuits and other similar requirements. More details are again found in the “Draft ETR”.
PT9V
The draft of the directive was ready in the spring of 1991 and at the same time the EC commission asked ETSI to make a plan for the standardisation work. This work was delegated to a project team, PT9V, under the ETSI sub-technical committee BT2, whose area is complex user installations. I was a member of PT9V.
Early October this work was so far that the scope of the real job – to write the standards – could be evaluated. Thus, ETSI could approve this next step and call for candidates to the project team PT22V, which should do this real job.in the course of 1992.
The future for ONP-LL
The first LLs shall – according to the directive – be available in January 1993, but no standard will be ready at this early time. The reason is that when the standard is ready from project team and BT2, it shall be approved in the technical committee BT and be submitted to a public hearing, before a referendum to approve it can be made.
This is necessarily a slow process as it lasts about one year from approval in BT or another corresponding technical committee to the issue of a standard.
Of larger interest are maybe the next subjects for standardisation under the ONP-LL directive. As to connection types the EC commission had already mentioned 34 and 140 Mbit/s. They could not be included in the first round. When they are standardised, connections through the synchronous digital hierarchy, SDH, will probably also be included. PT9V proposed that n*64 kbit/s connections with n = 2 to 30 should come under the directive, as such connections are of great interest for PABXes.
Comparison with Danish LL types
End 1991 Copenhagen Telephones offer the following types of LL for point-to-point connections without additional signalling:
• 2 and 4 wire analogue of local quality.
• 2 and 4 wire analogue of normal quality.
• 2 and 4 wire analogue of special quality.
• 4 wire analogue of broadcasting quality.
• 64 kbit/s digital with V.35 or V.36 interface.
• 2048 kbit/s digital structured with G.703 interface.
Local quality, for DC up to 15 kHz is only offered with both endpoints in the same exchange area and are not of interest in relation to the ONP-LL directive.
The next two types correspond to types in the ONP-LL directive. Probably their properties shall only be adjusted at a few points before they are in agreement with the new standards.
Broadcasting quality allows for transmission of 40 Hz to 15 kHz. These LLs are outside the types in the ONP-LL directive.
64 kbit/s digital LLs shall be available with another interface than today.
2048 kbit/s digital LLs shall be available as today. In these LLs the user has access to 1984 kbit/s as 64 kbit/s in a framing channel are outside his control. Unstructured LLs, in which the user has access to all 2048 kbit/s shall also be available.
Other directives concerning ONP
It was already said that LLs are only the first area where the telecommunication networks shall be opened according to ONP rules. Early in 1992 a plan is expected for use of these rules within the ISDN (Integrated Services Digital Network), and later directives can be expected for availability of IN (Intelligent Network services such as service 900, secretary services, facsimile distribution and voice-response services) under the ONP rules.
It means for the users that those applications of the telecommunication networks which were until now under the monopoly are opened for other service providers. Competition will be introduced on services and it will then turn out whether the rates of the monopolies for e.g. “call notification” are too high or if they are high because they are served by human operators. And whether the administrations or others can provide similar services using only technical means and therefore provide them cheaper – and still find customers!
-----
The work to prepare the ONP-LL standards fell in three parts. A short meeting in Sophia Antipolis end of June, 14 days in the end of July and most of September and October. I used of course the opportunity to have the family on vacation during a week of the latter period. Through ETSI I had rented a holiday apartment centrally placed between Antibes and Juan-les-Pins and it was rather small when we lived four persons in it. This time my family came by air. I had also this time spared some of my free days to the vacation and we made some fine trips. We took by train to Monaco, where the Oceanarium, the Botanical Gardens and the Caves made a great impression. The changing of guards was to us rather ridiculous compared with the Danish Royal Life Guards – and even the Tivoli Guards! Our oldest daughter was also shown the door in the Cathedral as her shorts were too short!
I had rented a car just as the year before, but this time only a Peugeot 205. It turned out to be large enough, however, to take us all four with our luggage to the airport the last day.
Cheap housing and cheap car rental. All in all I think Fischer and Lorenz earned well on having me as part of PT9V. Maybe so much that it compensated for their losses on my part in PT26!

Siemens HICOM


Early in 1992 Erik Andreasen called for me to give me a top secret assignment. I were to be project leader for an evaluation of whether Siemens HICOM PABX should be included in CTBD’s program of large PABXes as an alternative to L. M. Ericsson’s MD110. An engineer had been employed to take care of the details. He was so secret that he had no office at CTBD but at Siemens. Tage Fox Maule was also assigned to the project.
The three of us took part in meetings and were on a course in Munich. I was also in Berlin on a one-day trip to discuss the connection of “customer links” (2 Mbit/s PCM connections) to HICOM.
An exchange came to Denmark and was installed in the Lyngby exchange building. The two others studied it in depth. Several others were called in to work on sales preparation and sales opportunities. The discussions about its price were particularly difficult. Siemens failed to understand that we wanted an exchange without their special facilities at a price at the MD110 level, such that it could compete. They rather found that if we only sold it with all their bells and whistles it should be possible to ask a much higher price. But – the market would then also be very little.
During the autumn it was decided that CTBD and the corresponding division of Jutland Telephones should be combined under the head of the Jutland division, Søren Jensen. The two divisions had invaded the area of the other part and started a mutual price war. This was devastating for the earnings of Tele Danmark. The final blow was when a large order for PABXes to the public administration in Copenhagen went to the Jutland division at rock bottom prices for their Meridian exchange.
The combination led of course to a re-evaluation of the PABX offerings. It led to the marketing by Tele Danmark Business Division (TDBD) of the two systems they already had on their respective programmes. Siemens was notified that the work on HICOM would stop.
The liberalisation was already a couple of years old and Siemens choose to market their system directly. Thus they have probably learned that prices in Denmark were very low compared with several other markets and that nobody wanted to pay much for a PABX with basic facilities. The technical development and the competition were hard for all suppliers. From a price to the end user of about 6000 kr. (USD 1000) per extension in 1980 (a price which made no supplier rich) through about half of this in the late 1980es, the prices decreased to 1500 kr. (USD 250) per extension in the middle of the 1990es at the same time as there was still more individual adaptation of the systems. It didn’t really fit but that is what the suppliers, i.a. TDBD had to live with!

The N4H group


In 1992 the N4H group was started with Knud Solholt from Jutland Telephones as chairman. He was also the Danish representative in the ETSI technical committee TE5. This committee was responsible for ETS 300 001 or NET4, the rules for terminals on the analogue switched telephone network (or PSTN, Public Switched Telephone Network). NET4 from 1991 was a catalogue over the different rules in Europe, as national deviations were OK then. The differences were only removed when a change was made to TBRs (Technical Basis for Regulation), and TE5 had started a project to harmonise these rules. I became part of the N4H group because I had been part of the ETSI project team PT26, NET4 editing group, in the autumn of 1990.
The N4H group should evaluate the papers from TE5 from a Danish viewpoint. The name meant of course NET4 Harmonisation. The papers started with a discussion of which requirements to the terminals it was possible to harmonise and then went on to what the harmonised values should be.
We took most of the European discussion very quietly. We let the big network operators carry the burden and only occasionally our discussions made Solholt try his influence in TE5.
One of the essential questions was whether the loop resistance could be harmonised? The loop resistance is the DC resistance a terminal in its off-hook state presents to the network, such that the network perceives the terminal as busy. The Danish requirement in NET4 was that below 16 mA loop current the voltage between the terminal connec-tion points should be below 6,4 V, and above 16 mA the DC resistance between the connection points had to be less than 400 Ohm. This value or a lower one could very well be harmonised in our opinion. Although it might be a tightening related to what was required in Denmark if it was reduced to e.g. 300 Ohm, there would be no problems getting terminals complying with the requirement if it was a harmonised value.
However, an item quite impossible to harmonise was the French requirement that a terminal had to limit the loop current to a maximum of 60 mA. They stated it was necessary as a higher current from the line circuits in the exchange would saturate them and cause a lower quality of the transmission if the current was higher. Everybody else found that this was a problem handled by the exchanges themselves, making the terminals responsible could only be considered a technical trade barrier.
The discussion in ETSI continued for a long time after our group had changed into the N4G group with the task to revise the Danish part of NET4. Only in November 1994 the ETR 075 was issued with the result of the efforts by ETSI to harmonise NET4.

The DECT group


Another group in which I participated from 1992 was the DECT group. DECT is Digital European Cordless Telephony and is a technique to allow many users to have each his cordless telephone network in use within a small area. DECT is not an independent network like the mobile telephone network, each DECT system works under a public or private (PABX) exchange or network of exchanges in the PSTN. It could be used e.g. by the owner of a PABX to give its extensions local mobility unhampered by a neighbour giving the same opportunity to his employees.
LME (L. M. Ericsson) had done early work in this area and had developed and had in test operation several systems according to the same principles, but at frequencies around 900 MHz. The standards were now so far that a range of frequencies around 1900 MHz had been reserved and various protocol questions had been settled. LME could supply systems according to the latest version of the standards and the Danish telecom administrations would like to test them in different applications.
An arrangement was made for equipment for four tests, one at each administration. In Jutland it was set up under a public exchange in Aalborg as an alternative to the copper pairs in the subscriber cables, In Southern Jutland it was set up under PABXes in the frontier town Padborg to give local mobility to the personnel of the transportation firms. On Funen it was set up under the PABX in the headquarters of the administration to give mobility to the employees. In the Copenhagen area it was set up in a shopping centre in Hillerød under the public exchange. The handsets should work in parallel with people’s PSTN telephones, such that the users could have their telephones with them when they were shopping.
The DECT group prepared the tests and should evaluate the result. It was indeed started but died out due to the changes in the organisation (i.a. the fusion of the business divisions). It is my impression that there were never many customers to the application in Hillerød. Maybe because most people do not want to be called on the telephone when they are shopping – and the others use a proper mobile telephone, the use of which was not limited to one shopping centre!

Øberg leaves CTBD


Already before the fusion of the business divisions the “war” between them had caused heavy discussions within each administration. This also led to Øberg taking his leave. Some years later when I met him in Horsens he said it was a disagreement about what CTBD should do between him and the chairman of the board, the CEO of CT Jørgen Lindegaard, which had caused him to quit his job.
He was not unemployed for long. He told me that he was called by the CEO of Alcatel Kirk with an offer to replace him. It suited him better to become CEO of S. C. Sørensen in Randers (I think it was a family business). Thus Øberg was employed here and travelled between their sites in Ballerup and Horsens until the Alcatel operation in Horsens was closed down. Their PABX production could not any more compete with the large players on the market and Alcatel could in stead import PABXes to the Danish market.
Other parts of the production of Alcatel could also not continue, but they were bought by the employees and continued under the old name Kirk. It was the production and further development of transducers, i.e. microphone and telephone capsules for fixed and mobile telephones, a large export item as the employees were and are experts in this area. And it was the production and further development of telephone sets sold both in Denmark and abroad. Alcatel concentrated on the production of space electronics in Ballerup and the sales of telecom equipment from abroad.
I also left the fused TDBD, but not on my own initiative and far less dramatically. Small PABXes up to 40 extensions should now be sold by the main companies, so many employees from TDBD were transferred to CT and Jutland Telephones. I was one of these although I never had worked with such small PABXes while in CTBD. I came to the CT Telecom Division in the spring of 1993.

ETSI STC BTC2


By January 1st 1992 those of the employees of CTBD who had from old times been em-ployed in line with public servants, should decide whether they would be employed on a contract in CTBD or return to “big CT”. I.a. Niels Kristiansen, who was in the PABX group, chose to return. He then had to leave the PABX group and I proposed – and got approved – that he was replaced by another CTBD employee, Torben Paaske.
Niels also had to leave the work in ETSI STC BTC2 (or BT2 as it was called then) where he participated along with Jørgen Hess Kristensen from JTBD, Jutland Telephone Business Division, (who was the Danish spokesman). I suggested that I should replace Niels and this was accepted. I participated in meetings in Munich May 1992, in Bath October 1992, Sophia Antipolis April 1993 and Oslo September 1993.
Munich is the home town of Siemens and CTBD had started discussions about the in-troduction of HICOM in Denmark with them a few months earlier. This should be kept a secret for our competitors in JTBD, so I kept mum towards Jørgen. Well, it corre-sponded very much to what I was used to from Philips. We were always very friendly towards people from our competition but kept mum about company secrets. And so did the others!
When I received the flight ticket, I also got a telecard (to use in pay phones) with a value of DEM (Deutsche Mark) 5 from Lufthansa. It had a Danish text urging to use the card during the trip. I did so, but during the walk to the meeting place I passed a shop on a corner of the main station advertising the buying and selling of telecards. This I would of course test and every price from DEM 1 upwards would be a profit! Thus, when the card had been used I visited the shop, the card was set in a reader (to check the remain-ing charge and to ensure it was a real telecard) while I muttered that a German telecard with Danish text had to be a rarity. Well, they offered 2 DEM for it which I gladly re-ceived.
The German Telecom, who had arranged the meeting, invited all participants on an eve-ning trip to Garmisch-Partenkirchen, where we had a good meal and enjoyed local per-formers – and a performance of some of us, who showed their muscle power by compet-ing about who could hold a filled beer glass in an outstretched hand for the longest time.
The meeting in Bath was called by British Telecom. I used the opportunity to postpone my trip home to Saturday, such that I could be a tourist Friday afternoon. Going by bus around in the town, visiting the Roman Baths and the Cathedral, the latter with i.a. the epitaph from one great actor to the other, de Garrick writing over James Quin with the potent final words: “Whate,er thy strength of body, force or thought; In nature’s happi-est mold however cast; To this complexion thou must come at last”.
Several years later I found these names in the book by Knud H. Thomsen “Røverne I Skotland” (The Scottish Robbers). But back in 1992 I had borrowed the book “Sarum” at the public library. This book is about Salisbury from the Ice Age to the present. This town and Stonehenge are not far from Bath and Bath is mentioned at about the time when the Roman Baths were built.
At the meeting in Oslo I was accompanied by my wife, Jytte. We made it a small holi-day, first a couple of days in the Swedish archipelago north of Gothenburg and after the meeting a Friday afternoon as tourists (which my wife had also been every day during the meetings). The meeting was well attended from Denmark, with 4 participants, i.a. Bent Qvist from Copenhagen Telephones just like me. This was because the last items in the standard for analogue leased lines should be decided at the meeting.
Friday Jytte and I made two trips to the woods North of Oslo. First we went by rail to a place with a beautiful lake which we circled, then by another railway to the Frogner hills. We made again a long walk and enjoyed then hot chocolate in a restaurant with a view to Oslo where we saw the boat to Copenhagen sail out.
Spring 1993, the Rouban Group
During the meeting of BTC2 in April 1993 I became chairman for a small sub-sub-group.
BTC2 was busy with standards for leased lines, the task I had helped planning in the autumn of 1991. The job was done in project team PT22V and it was then discussed in working party WPF of BTC2. My place was in WPA on transmission, but I got mixed up in the plenary discussions of the results of WPF. In any way the requirements to a terminal on an analogue leased line as to the power it might transmit outside the speech band was thoroughly discussed at the meeting in Sophia Antipolis.
The problem was that while the requirements to the connection, which the telecom ad-ministrations should comply with, were comprehensive in order to ensure that the cus-tomers got a connection of a certain quality (as regards attenuation, distortion, delay, noise etc.), the requirements to the terminals were governed by the Terminal Directive. It lists 7 essential requirements to which a terminal must comply, viz. 1) user safety, 2) safety for the telecom personnel, 3) EMC (no disturbances to and from other apparatus), 4) no harm to the network, 5) a reasonable use of radio frequencies, 6) cooperation with the network to make, keep and release connections and 7) cooperation with other termi-nals via the network in justified cases. Requirements 1), 2) and 3) shall be complied with through other standards and 5), 6) and 7) were irrelevant for terminals on leased lines (no radio transmission, the connection exists all the time and can not be changed from the terminal and 7) was to be considered for speech transmission only). This left only 4), no harm to the network, and what did that mean?
There were fanatics (“liberals”) who wanted to limit this to mean only physical damage, such as when a high voltage from a terminal could make network equipment fail. We obtained however agreement that it should also mean that a terminal was not allowed to send signals disturbing other users of the telecom networks. Such was the stance in April 1993. I had made a proposal ensuring that the signals sent could not via crosstalk (i.e. coupling to other wire pairs in the cable) disturb others. This was however not enough. During the meeting it was emphasised especially by the participants from Brit-ish Telecom that transmission systems where connections are multiplexed on a common path did not isolate the channels completely from each other. On a carrier system the channels are placed besides each other within a range of frequencies. If a channel sends too much power at the frequencies of a neighbouring channel it will disturb. The filters for each channel are not ideal, partly as a question of price, partly as a question of the technique. If a filter cuts off unnecessary frequencies too sharply it will distort the wanted frequencies close to the edge. All this was a quite justified viewpoint.
The discussion made the Chairman of BTC2, John Horrocks from British Telecom, pro-posed a group formed to present a proposal for requirements at the WPF meeting in July and he asked me at the meeting if I would chair this group.
I made a long pause. Yes, I would like to, but I had recently come to CTTD, Copenha-gen Telephones Telecom Division, (when CTBD and JTBD were joined into TDBD and small PABXes were moved to CTTD), and I had sensed the attitude to the work in ETSI. Thus it was clear for me that the job as chairman had to be handled without trav-elling. It was also only for a short time. The thoughts ran through my head until I saw a way and accepted the job. I collected names and fax numbers of the interested BTC2 members.
As soon as I was back home I described my ideas in a fax to the interested people. At the same time I coined the name The Rouban Group for “Requirements to OUt-BAnd Noise”. The idea was to send my proposal as far as it went by fax on May 1st, June 1st and July 1st, and then the interested parties could fax their comments by the 15th of each month. The schedule was to call for proposals in May, to make a draft by June and a final draft in July. There would just be time to get a yes or no to this before the WPF meeting in the end of July.
It was carried out as proposed. There was a good proposal from Holland and it was re-written to the standard formulation of ETSI. There were some comments but nothing disastrous in June, so it was largely this proposal which went out in July. It was unani-mously accepted (at least nobody had comments to the changes from June to July). Brit-ish Telecom even wrote that they were very satisfied and could support the proposal 100%. It also passed through the WPF meeting without amendments except for a few minor adjustments. One was that L. M. Ericsson rather late had observed that the im-pedance over which the power was measured was not the correct one outside the speech band. I had confirmed this in a comment to the final proposal and the change was intro-duced.
The problem was that inside the speech band a fixed impedance is used equal to the im-pedance at 1020 Hz, although the impedance changes over the band. When one moves outside the speech band and as here even up to 1 MHz the impedance deviates however too much from the impedance at 1020 Hz to keep it at this value.
The results from the Rouban-group were included in the requirements to terminals for analogue leased lines according to the ONP-LL directive, TBR15 and TBR17.
It is my impression that much standardisation work could be done as in the Rouban-group by contact via fax or e-mail, when people have first met and got to know each other. It is also my impression that the few meetings are well prepared by these means so everybody is fully oriented about all viewpoints before them.
The N4G-group
During 1993 the work on NET4 continued in the N4G-group. The job was to revise the Danish part of NET4 (or ETS 300001) to be part of a new version of the standard.
The first version had an exception for PABXes and there was in stead a reference to national regulations for them. These words would be removed, but NET4 had on several points values different from those in the Danish requirements to PABXes.
The PABX-group was by the way busy revising part 3 of these technical requirements, which treated their connection to analogue subscriber lines in the switched telephone network. I was editor on this task. We were ready a couple of months before the N4G-group was ready, but the Telecommunication Supervisory Board decided that the new version should not be issued while waiting for NET4. All these efforts for nothing, but the work was of course valuable for the work in the N4G-group.
One of the great differences of opinion within the group related to the leakage resistance of the terminal to earth. Circular 2 (requirements of the network to subscriber terminals in general on the switched network) said that the subscriber installation as a whole was allowed to have a minimum of 10 MΩ from each wire to earth. The first issue of NET4 required a minimum of 100 MΩ from each wire to earth for each terminal (several might be connected in parallel). At an early moment we had seen in the PABX-group that this was a too small value for a PABX when it should measure the polarity over the wires (used to check if the subscriber line was busy in the public exchange or not, in order to release the line as soon as an external caller had gone on hook). The PABX would require at least 10 MΩ, and as there might be more terminals in parallel con-nected to the NTP (Network Terminating Point) Circular 2 had to allow a correspond-ingly smaller leakage resistance of the total subscriber installation. Considering that the requirement to the subscriber cable was a minimum of 100kΩ from each wire to earth or 100 times less than required in NET4 of the subscriber terminal there was an ample margin to exploit. We had already in 1989 discussed this point with the technical com-mittee of the telephone administrations and received an accept of a minimum value in Circular 2 for the whole subscriber installation of 3 MΩ from each wire to earth. We PABX specialists used this value repeatedly in the discussions in the N4G-group, as this was our basis for requiring the value in NET4 changed from 100 to 10 MΩ, such that NET4 could also apply to PABXes.
We did not succeed completely. The revised text measures leakage resistance between short-circuited line terminal points, i.e. in the worst case half as much as from each point to earth, and this value was set to 10 MΩ in general. A considerable reduction compared with the first version where the same value would have been 50 MΩ. How-ever, the text was made acceptable due to an exception for PABXes. They were allowed a leakage resistance between the short-circuited terminal points and earth of minimum 5 MΩ.
This implied of course that the term PABX had to be defined in NET4. NET4 already had requirements to terminals connected in series, defined as terminals with one port towards the network and one towards a NET4 terminal, where the second terminal only receives power for its operation from the first terminal. We worked much on the defini-tion of a PABX and got as the result that a PABX was any other terminal with these two types of ports. Thus, if there were only two ports it was a requirement that the terminal on the second port was at least partly fed locally and if there were more than two ports it was by definition always a PABX.
The work was ready in the end of 1993 so it could be notified within the EC and could be sent to ETSI. It was, however, not included in the April 1994 version of NET4. It was only included in 1995 and the Telecommunication Supervisory Board issued a Dan-ish version of NET4, which only included the common text with the Danish values and the special Danish amendments such as the definition of a PABX.
The PANT-group
In 1994 I became part of a group preparing a buyer’s specification for an NT, Network Terminal, for ISDN PRA (Integrated Services Digital Network, Primary Rate Access), i.e. the unit installed by Tele Danmark (TDC) at customer premises to provide the inter-face to a 30 channel connection on a 2 Mbit/s subscriber line. The specification should of course be based on the standards from ETSI, and they covered the digital line, errors on it as well as the T interface, which is the interface to the customer’s equipment. The group was established by IU-A, the Infrastructure Committee, Access, of TDC.
But as always there were options, possibilities for choice, in the standards and our task was to choose which of these should apply to the NTs bought by TDC.
Chairman of the group was Ole Philbert from Jutland. Members were Erik Bisgaard and Poul Erik Pedersen from Jutland and Otto Kimer and me from Copenhagen.
The purpose of such a specification is of course to make common rules for the behav-iour of the equipment, such that we were not tied to use only NTs from LME under LME exchanges, or NTs from Siemens under the exchanges they had supplied. It should also be possible to buy NTs from a third supplier and use them under both types of ex-changes. The more the different network elements can be independently bought, the better prices can be obtained by TDC!
We had many meetings discussing the possible choices and their consequences. With my background I insisted from the start that the NT we should specify ought to be us-able also on “subscriber links”, 2 Mbit/s digital lines with switched telephony used as subscriber lines to digital PABXes, and on 2 Mbit/s structured leased lines, which we should provide as a part of our obligations according to the EC directive about leased lines (ONP-LL) and where I had participated in the planning of the standardisation work in ETSI PT9V during the autumn 1991. It turned out that the various requirements fitted together, such that this could be realised.
It also turned out that I got the job of collecting the contributions and reformulate them to the required application specifications to the ETSI standards. There were three stan-dards, one for general conditions (ambience etc.), one for the T-interface and one for the digital line.
The job was finished in January 1994 and the group was disbanded a couple of months later. In July 1994 TDC then issued their technical specification TDK-TS 900213 with our specifications.
I got mixed up in the use of the specifications later. Poul Erik sent me compliance statements, i.e. confirmations that the requirements were complied with, from LME and Siemens for evaluation. I remember that Siemens did not comply with one of the re-quirements, that the NT towards the customer should set the bits S(a5) and S(a6) used on the line to the value 1. Siemens had chosen to set them at 0. Back to the ETSI speci-fication for the T-interface: it required that the terminal should ignore these bit. We had made the requirement mandatory, but could accept the choice made by Siemens. If we later had to revise the specification we would probably drop our requirement.
Later while in the network division (after the old administrations had been replaced by functional divisions in TDC) my boss, Martin Gregersen, came with a data sheet for such an NT from Fujitsu. I said to him that he should ask Fujitsu to make a compliance statement according to TDK-TS 900213 as a basis for further discussions. I do not know whether this led to our buying these NTs.
Finally, in 1997, Poul Erik asked me if I still had the files from the past. They had been included in INFR, ISDN Functional Requirements, and these should now become TDK-TSes. Thus it would be good to have the original files. Luckily I had always made back-ups on floppies and had brought them with me from CTBD via CTTD to the network division. Thus I could go back in spite of the change of PC and operating system and send them to him. Another piece of luck was of course that the floppies could still be read although I had changed PC!
The PSTN-group
In 1994 I became part of a group specifying the monopoly interface in the switched ana-logue telephone network. This was probably another consequence of my work in ETSI during the autumn of 1990.
Our chairman was Erik Bisgaard from Jutland. The group was established by IU-A, the Infrastructure Committee – Access network.
The monopoly interface or - as it was later called - the NTP, Network Terminating Point, is the point where the TDC cable meets the subscriber installation. The liberalisa-tion meant that not only were the terminals, but also the whole installation at the sub-scriber premises, outside the TDC monopoly in making telephony connections. A de-scription was required of what TDC delivered in the NTP and of what we delivered be-tween two NTPs, corresponding to the network interface and the connection properties in the specifications for leased lines.
We had a good starting point in the old Circular 2 (from the 1970es), which had a sec-tion dealing with the subscriber installation. There was i.a. a requirement for a minimum loop current of 15,7 mA when the NTP was terminated in 600 Ω, and the distribution of the 600 Ω was described (400 Ω in the final terminal, 200 Ω in the subscriber installa-tion including serially connected terminals). This had to be modified as the requirements to terminals were changed since the 1970es (the old Circular 27), e.g. the requirement to the final terminal at loop currents below 16 mA was not any longer a maximum of 400 Ω, but a maximum of 6,4 V over its terminating points, and this had to be reflected in the properties of the NTP. I emphasised that as far as possible we should specify the requirements to be in line with Circular 2.
At the same time IU-T, the Infrastructure Committee – Transport Network, was busy on a similar job. They rather looked at how little TDC could provide in the NTP and would base their efforts at the work in ETSI to make a common European set of requirements to telecommunication terminals in TBR21, i.e. requirements ensuring that a terminal might be used throughout Europe. Their attitude was that the network should not prom-ise better properties than just required to serve a TBR21 terminal.
My standpoint – to which my group adhered – was that it was not the network which had to adapt to TBR21, but the TBR21 terminal which had to adapt to the network. Cir-cular 2 was the Danish starting point. It said e.g. 600 Ω terminal resistance in the NTP. If another network operator required 500 Ω terminal resistance and this became part of TBR21 it would be OK for us, but should not change our specifications for the network. If on the other hand it was sufficient for all other European network operators to require a loop resistance of 800 Ω, we should require of TBR21 that its requirement was our 600 Ω. Thus, our group insisted on maintaining the 600 Ω loop resistance over the NTP, while the other group would adhere to the requirement of maximum 500 Ω loop resis-tance in the terminal according to TBR21, and then we should not promise that it would be perceived as a loop (off hook) if there were more than 500 Ω loop resistance over the NTP.
This ended up, as one can guess, with two specifications with slightly different contents. The shape of our draft became the final one, but it was decided in IU that the figures from IU-T should be used, which they were. I got angry during this, when it was re-ported that the chairman of IU-T, Ole Jahn Svinding from Jutland, had said in an IU meeting that there was agreement about the figures from IU-T. This was blatantly incor-rect. But in the final discussions Erik Bisgaard was under a large pressure, and therefore the figures from IU-T were included in the final specification, issued as TDK-TS 900221.
Of course I could not help laughing later when the figures in TBR21 were modified (and TBR21 was by the way not finally approved as late as in the spring of 1997), meaning that there is no more a reference for the figures in TDK-TS 900221!
ETSI STC TE5
A consequence of the work in the N4G-group and the PSTN-group was that I also be-came part of ETSI STC TE5. However, I took only part in one meeting, in Mainz in October 1994, invited by the German Telecom.
At this time TE5 was busy finishing TBR21, a common European NET4. It applied to non-voice equipment, a corresponding TBR37 should be issued for telephone sets with TBR38 valid for the speech transmission itself.
There had already been lengthy discussions of which requirements to the terminal could be called essential and how the chosen requirements should be worded. Each country had expressed itself in its own way and was of the opinion that their wording of the re-quirements was the best one. In Denmark we had e.g. a requirement for a certain real part of the ringing impedance plus a certain numerical value of it. Germany had a re-quirement for a certain capacitive value of the impedance plus a numerical value. The requirement in TBR21 might e.g. be the common window valid for all countries – if such a window existed!
If it had been my baby I had insisted on first defining what was essential and then find-ing possible common windows on these points. Then it was not necessary to discuss those items any more.
A significant difference from NET4 was that TBR21 should only apply to a single ter-minal connected directly to the NTP (Network Terminating Point). The requirements in the Danish part of NET4 were based on having up to 200 Ω and 1,5 dB attenuation be-tween the NTP and the final terminal. Thus, the network will deliver at least 40 V AC as ringing voltage over the minimum ringing impedance connected in the NTP, while the requirement to the final terminal is that it works on 35 V AC ringing voltage. The re-quirements are also based on having more terminals in parallel, so the ringing imped-ance of each terminal must be a multiple of the minimum ringing impedance required by the network. All requirements in NET4 had to be adjusted in this way to apply in NTP for a single terminal before the common windows could be defined.
There would of course be several points left where it was not possible to find a common window. The discussions could then concentrate on these items.
I had arrived so late in the group that it would be impossible to introduce a new method for deciding the values. The efforts of us Danes was therefore only to see to it that TBR21 did not contain items which we could not comply with and for the rest try to help finding acceptable compromises when we had no special requirements to fight for. Typical was the French insistence on a requirement that the terminal should limit the loop current to 60 mA. All others had to accept this and worked then in unison to find out how the requirement should be worded for loop currents under 60 mA in such a way that it would still be possible to produce the terminals. The compromise found was that the terminal had to allow a loop current of at least 40 mA on the shortest loops.
Telecom invited on a trip one evening to a vineyard as it was in the period of wine har-vesting. We learned about the production of wine and it was a surprise for me that the first fermentation only takes about 14 days. During this time it is dangerous to enter the fermentation room due to the amount of carbon dioxide produced. The vineyard ran a restaurant in the village where we had ham and sausages – and an opportunity to taste their different wines!
Jørgen Lindegaard moves to GN Great Northern
In the autumn of 1995 it became impossible for Jørgen Lindegaard to continue as CEO of Copenhagen Telephones. His chemistry did not fit in with that of the CEO of the Tele Danmark concern, Hans Würtzen. Jørgen decided to leave Tele Danmark for the job as CEO of GN Great Northern, the oldest telecommunication company of Denmark, dating back to the 1850es when it was formed by C. F. Tietgen as a telegraph company.
A large farewell reception was held and I came of course too to say good by and good luck to Jørgen. He surprised me by not only accepting the good wishes but by asking me immediately whether I had a place in the new organisation due to start on January 1st, 1996, and which meant the dissociation of the old telephone administrations. Indeed I had, and he was glad to hear this. I had not expected him to think about other things than his own farewell to Copenhagen Telephones on this occasion, but was happily surprised.
Changes in the Telecommunication Supervisory Board
There was another change this autumn. In the TSB, Telecommunication Supervisory Board, Ernst Hansen retired when he became 60 years and there was a farewell recep-tion to wish him good luck.
A successor was called for via advertising and I must admit I felt myself qualified for this job. This was confirmed when Poul Tækker who I knew from i.a. the N4G-group phoned and asked if I had seen the advertisement? Indeed I had but I would not apply. I did not feel on the right shelf in the new organisation of Tele Danmark (I came further away from the exchanges), but had decided to stay. I did not say this to Poul Tækker but answered in stead that I did not expect them to choose a person of 62 years and soon 63 to replace Ernst Hansen who had left at 60!
The TSB found a replacement in Helene Jensen who came from the Civil Protection Agency. We had a fine cooperation when I became chairman for the TAFK-group, see below.
Legislation on Telecommunications
During these years a lot happened within the telecommunication field concerning liber-alisations of not only the terminals but also the whole infrastructure, aiming at having a fully free telecommunication market in the EC from January 1st, 1998 (except in the most backward countries). The EC issued directives and they should be followed by legislation in the EC countries. In Denmark the liberalisation was even hastened so the area was free already in the middle of 1996. However, the necessary legislation was only ready during the first half of 1997. It caused several contributions to journals about this topic. I found that one of them, in “IT Standard News”, deserved a comment and sent the following to the journal, where it was published early in 1996:
“A few comments to the article “The EC commission changes its view on the need for legislation concerning telecommunications” in IT Standard News, October 1996.
About the background Birger Jacobsen (BJ) writes that the philosophy behind the rules for terminals was the necessity “to protect the network itself against damage caused by the terminals” and that this has caused a comprehensive set of rules. The need for pro-tection itself has, however, not implied a lot of rules, see e.g. the draft regulations such as TBR15 for terminals on leased lines under the ONP-directive (Open Network Provi-sion) which may only be justified by this requirement. The many rules will only come into play due to the essential requirement for cooperation with the network to make, uphold, release etc. connections. But these rules are also not very comprehensive as they must be limited to the mentioned functions. Thus, no requirements can be made as to the correct function of supplementary services in connection with a type approval.
Within the present legislation BJ mentions the rules concerning voltages in the Low Voltage Directive (LVD) and that “the Terminal Directive (TD) had therefore to sup-plement the LVD as to other voltages”. This is regrettably not correct. The safety rules in the TD (91/263/EEC) may only be complied with through a reference to the LVD. This means that a terminal supplied from less than 75 Volts DC shall meet no safety rules due to the TD. Quite reasonable, neither users nor personnel from the telecom op-erators will be killed if such a voltage is leaked. However, there are some non-electrical requirements which are not covered and must not be covered, such as a requirement that the sound level in a telephone must not be so high that the hearing ability is spoiled. When the TD is revised special conditions within telecommunications ought to be al-lowed treated in regulations based on the TD, such that this requirement could be rein-stated.
During his treatment of the history BJ mentions as an example of damage to the network that a terminal “makes the whole network die”. This is a good illustration of the need for a definition of the content of the essential requirements. It is of course also damage when a part of the network dies, or if the terminal sends signals with such a power that they spill over to other connections. At this point the Danish TSB has by the way been more liberal than the EC, as it has removed the rules for terminals on digital leased lines. The network terminal from the operator – to which the user terminal is connected – protects the rest of the network, and if it is damaged it will not influence other connec-tions. With the introduction of ONP for leased lines rules for these terminals appear again, as the network terminal is perceived to be a (protected) part of the network.
BJ also writes that one might (maybe) conclude that “it is more important to keep the telecommunication network alive than to keep the users alive”, referring to the fact that safety requirements, EMC requirements and requirements to the network are only based on voluntary standards. BJ forgets that with the references in the LVD and the EMC-Directive and with the references in the ONP-Directive to what the operators shall de-liver, these voluntary standards are made obligatory.
Finally, one can to the four bullets about the proposal from the EC Commission add:
• If the requirements are limited to damage on people and network while the co-operation with the network is based on an agreement between user and operator one might open for a situation where a terminal can not be used on more net-works. Should a situation again be made in which each operator (Tele Danmark, Telia etc.) supplies its own terminal program together with the subscription?
• One can not obtain interoperability and harmonisation by legislation for the net-works only. The terminals must also comply with certain requirements in order to cooperate with the network. These requirements should be the same for all terminals for a certain service, independent of the network operator.
• It must remain essential that terminals on the fixed network (for both switched and leased lines) do not disturb other connections. To increase the level indis-criminately can not be tolerated.
• The terminals must thus continue to comply with certain requirements, but it can be better to base their access to the market on declarations from the producers about this as it is already done concerning safety and EMC.
• It is correct to maintain that all use of radio frequencies must be regulated.”
The TAFK-group
I became a member of this group when it was started in 1994 to revise the technical regulations for terminals for analogue leased lines.
As a basis we had the Publication 12/90 from the TSB, which had as an appendix the old Circular 12 from the tele-administrations. These administrations had also made a proposal for a revision, which had not been approved by the TSB.
The group was established according to the new procedures in the TSB, i.e. as a group formed by the TSB for which Tele Danmark provided chairman and secretary plus pos-sibly other members free of charge. Other organisations might also provide members.
Leo B. Rasmussen from Jutland became chairman and Arne Crown Rasmussen from Funen became secretary. Members at large were Ebbe Christensen from DELTA Elec-tronics Test, Ole Skjoldstrand from the Society of Electronic Factories and me.
The starting point was that ETSI was standardising some types of analogue leased lines under the ONP-LL Directive (Open Network Provision – Leased Lines) which would replace some of the types provided up to now by Tele Danmark, and which we had to provide (they belonged to the minimum set of leased lines along with digital 64 kbit/s and 2 Mbit/s lines which had to be offered throughout Europe). As to these types the new technical regulations for the terminals should just refer to the CTR’s to be issued by the EC Commission based on the TBR’s from ETSI (Common Technical Regulations, resp. Technical Basis for Regulation). I had been a part of the planning within ETSI to prepare these regulations as a member of PT9V in the autumn of 1991.
There were four types of analogue leased lines for which this applied viz. 2-wire and 4-wire lines of normal and special quality. There were, however, only two TBR’s, resp. TBR15 for 2-wire and TBR17 for 4-wire terminals. The difference in quality came in what the network operators should deliver (e.g. less variation of delay with frequency for lines of special quality).
But in addition to these types Tele Danmark also offered lines with signal conversion, with local quality and with broadcast quality. For these types of lines we had to go in detail with the requirements to the connection of terminals in the new technical regula-tions. At the same time we should adhere to the Terminal Directive from the EC, which said that it was only essential requirements which should be complied with in order to get a type approval for a terminal. Considering leased lines this meant that it was only possible to require such properties that the telecommunication networks were not harmed.
This caused many discussions in the group. My starting point was that there should be as wide limits for the terminals as possible, while Leo rather wanted more limitations. As to lines of local quality which were supplied as wire pairs between the customer ad-dresses (and were therefore normally limited to have both end points within the same exchange area) Tele Danmark specified that they were for transmission up to 15 kHz. In real life wire pairs can be used at much higher frequencies, e.g for ISDN2 connections using power up to about 50 kHz (and think of the DSL lines during the 2000es). My viewpoint was that customers should be allowed to use the lines at the highest frequen-cies they wanted, e.g. to an ISDN2 terminal under a PABX but at another address. It should only be required that they should at each frequency send at such a level that cross talk could not disturb neighbouring circuits. It should not be required to select wire pairs for this type of lines.
In the IEEE Communications Magazine there had been an article (Chen and Waring, May 1994) stating that the crosstalk attenuation decreased by 15 dB per decade, i.e. with 15 dB from 10 to 100 kHz and again with 15 dB from 100 kHz to 1 MHz. The labora-tory of Jutland Telephones could confirm that this was also the case in Denmark (it had been measured when ISDN was introduced). As a consequence this value was stated in the technical regulation as valid for the range from 15 kHz to 1 MHz.
In the spring of 1995 we could send our draft to the TSB and during the summer it was published for comments. The TSB had divided it in 3: TB95069 referring to TBR15, TB95070 referring to TBR17 and TB95071 for terminals on the special Danish lines. Comments were received from Tele Danmark to the latter and then nothing happened for a long time.
The reason was internal changes in the TSB. Ernst Hansen retired and Helene Jensen replaced him. She came from outside the TSB and had to work herself in first.
In the meanwhile other things happened. The EC Commission had decided that the only case in which the essential requirement for cooperation between terminals via the public network should be applied in connection with analogue terminals was for speech trans-mission between normal telephone sets directly connected to the NTP’s (Network Ter-minating Points). As a consequence the requirements to transmission and to the inter-face towards the extension, such that transmission from a normal telephone was possi-ble, in the technical regulations for PABXes were to be discarded. We had in TB95071 referred to exactly these requirements when writing the rules for terminals for lines with signal conversion from and to the usual telephone signals (ringing voltage and loop in-terruptions) on 2-wire lines.
Thus the group was again activated in the autumn of 1996, but in the meantime the two others from Tele Danmark had moved to other jobs within the company and could not continue. I had then to continue as both chairman and secretary. The work to include the requirements directly in TB95071 was, however, rather easy. We had only one meeting and managed for the rest with a few faxes before we could in the middle of November send our draft to the TSB.
This did not end our job. It would only be finished when both a Danish and an English version of the technical regulation had been issued. We took this up with the TSB, which were also very responsive to our proposal that the final editing of TB95071 should follow our proposals closer than was done in the issue published for comment.
It is lightly ironical that the requirements added directly to the technical regulations were for a disappearing circuit type. Tele Danmark could not any more buy the signal converters necessary between the physical line and the transmission equipment. At the same time the transmission equipment always bundled the circuits in groups of 30 cir-cuits (PCM systems) which had to follow each other. This was not economically sound if you only needed one circuit. The signal converters could be replaced by interfaces to the LLSN (Leased Lines Services Network) which was being introduced and which had the required 2-wire interfaces. But between such interfaces there was still the require-ment that 30 circuits had to be provided in parallel.
Tele Danmark had other cards in its sleeve. Digital public exchanges were introduced in great numbers and their facilities could be used. Especially the Centrex offering, a coun-try-wide PABX functionality with no other equipment at the customer address than normal telephone sets, was useful. Using this offering as the main PABX and with cus-tomers having PABXes and normal sets under it the same functions could be provided as with leased lines with signal conversion. In addition the customer should not – as with leased lines – pay whether he used the line or not (leaving it also to Tele Danmark to somehow cover the cost of 29 maybe unused circuits). Now the connections were in the switched network and existed only when needed. When a connection was not in ac-tual use the different parts of it were available for other users.
The Part 8 group
Part 8 of the Danish requirements to PABXes treated their access to “customer links” or 2 Mbit/s PCM connections to the public exchanges. These are structured PCM systems, meaning that the bit stream is divided in 32 timeslots of 8 bit each, repeated 8000 times per second. Timeslot 0 always has a certain pattern, making it possible for the receiver to recognise it and find all timeslots (and bit). Timeslot 1 to 15 and 17 to 31 are the 30 speech channels in which speech (or tone signals) are coded according to the A-law (in Europe). Timeslot 16 is the signalling channel, in this case using CAS or Channel-Associated Signalling, meaning that every 16th timeslot 16 has a certain recognisable pattern, making it possible to associate the 15 remaining timeslots 16 with 2 speech channels each. Thus each speech channel is accompanied by 4 signalling bit, repeated every 2 ms.
The signalling bit are used for line signalling (seizure, release, answer), while the regis-ter signalling (forward signals like dialled digits, type of caller, backwards signals to call for next digit or caller type or to report if the called party is free or busy) were MFC (Multi Frequency Code) signals in the speech channels. Already in 1974 I had worked with this, although on analogue lines, for the DDI (Direct Dialling In) function in EBX 8000 for CTBD and later for the state network of PABXes. From the beginning this sig-nalling was very flexible, as proved when it was discussed with Siemens in 1992. Unlike other PABX suppliers they wanted to receive the signal for type of caller before the dialled digits, but this was no problem, if only they asked for it with a backward signal and then asked for the relevant digits afterwards.
In 1990 the PABX-group had issued Part 8 in a “provisional draft” form. It could not remain so. The reason was that there did not exist at that time a real possibility for test-ing whether the terminals complied with the requirements. In 1994 the standardisation of ISDN was so far that it was at least possible to make proper test specifications for the physical connection.
Thus, the TSB created in 1994 a group besides the PABX-group to make a revision of Part 8. It did of course mostly consist of people from the PABX-group. Tage Fox Maule from TDBD, Tele Danmark Business Division, became secretary and should edit the draft.
While the group was active changes occurred in what should be required from PABXes. It was laid down that requirements should only apply to the connection to the line to the public exchange and be limited to the basic functions. Transmission requirements should be scrapped, as the essential requirement for cooperation between terminals in relevant cases, article 4(g) in the Terminal Directive, should only apply to normal tele-phone sets for live speech when the sets were connected directly to the analogue Net-work Terminating Points. The TSB was also becoming stricter as to what a Technical Regulation might contain. It should in the future be limited to the requirements and the test arrangements making it possible to verify the requirements.
The total set of PABX requirements was therefore cut down considerably. There was still a Main Regulation (the old part 1 and 2 with abbreviations etc.) but part 3 about analogue connections to the switched telephone network was replaced by the rules in NET4, edited by the N4G-group. Part 4 about analogue connections to leased lines ex-isted still in 1997 (when I retired) but would then be replaced by the rules edited by the TAFK-group. Parts 5 and 6 about supplementary equipment and functions on analogue connections to the switched telephone network were scrapped. Part 7 about the connec-tion of analogue telephone sets on the extension side was also scrapped. Part 8 about connection to customer links should remain and be edited by this group. Part 9 about connection to digital leased lines was replaced by a reference to the Technical Regula-tions for terminals on such lines. Parts 10 and 12 about transmission in resp. analogue and digital PABXes were scrapped. Part 11 did never exist.
As said: Part 8 should remain. And considering that there was no Technical Specifica-tion from Tele Danmark for its delivery of customer links we did for a long time neglect the new guide lines from the TSB and kept an informative chapter with such a specifica-tion. We did of course also advise that Tele Danmark should issue such a product speci-fication.
For the rest the job was a careful examination of the old draft and the solving of all doubtful points. A funny aspect was the scope of requirements. No test laboratories were interested in making a set-up for automatic testing of terminals for customer links, the market for tests would be too small for it to pay off.
The outcome of this was that when we handed over our result to the TSB in October 1995 there was an informative chapter and the requirements were limited to cover only the physical access, as it could be tested in the set-up already used for test of ISDN30 (or primary rate). Whether the signalling worked towards the public network we left to be based on suppliers’ declarations.
The TSB did not like this – they had no authority to implement such a rule. They did neither like the informative chapter.
In 1996 the old telephone administrations had been reorganised into functional entities in the Tele Danmark concern and the work with specifications was arranged in a new way. One point was that John Hartrup in NIS (Network Infrastructure Switching) should make a technical Specification for customer links. He later moved to the regional or-ganisation for Copenhagen and the work was taken up by Jürgen Endres in NUS (N Development S) with assistance from Harly Østergaard Jensen, U (Development). They were both members of the new PABX-group under the TSB. This resulted in TDK-TS 900234 of May 1997.
This meant that the informative chapter in the Technical Regulation could be replaced by a reference to this TDK-TS. Thus, in June 1997 the TSB could finally issue TB97087 with access requirements for terminals on customer links. I did not see it before my re-tirement, so I do not know the result of the considerations concerning the scope of test-ing.
The BELA-group
A group was formed within Copenhagen Telephones in 1995 to describe limitations for electronic solutions in the access network. The group should compare such solutions with having a copper pair to the customer.
Copper wires can do much more than transmit the necessary frequencies for speech in the range 300 to 3400 Hz. Their attenuation does not increase steeply above this band, in stead it increases gradually with the square root of the frequency. If the attenuation is 3 dB at 1000 Hz, it will be 6 dB at 4000 Hz etc. Therefore one can work with rather high frequencies such as ISDN2 connections requiring transmission up to about 50 kHz or even ADSL (Asymmetrical Digital Subscriber Line), using the copper pair to trans-mit TV-signals at several MHz to the customer.
An electronic solution might be used in a place where there were no more copper pairs to avoid laying a new cable, by e.g. installing electronics at the two ends of the fully used cable to double its capacity by transmitting ISDN2 signals on its pairs and translate these signals to normal copper wire signals at each end of the “narrow” cable. But this will limit the transmission for the connections to the speech band.
This is not the only limitation. The copper wire is also used to feed the telephone instal-lation such that it will work even if there is no electrical power in the house. It might still be possible to feed two installations via a single pair, but if more connections shall use the same pair a local power supply is necessary including a battery if the installation shall be independent of the electrical power system. However, when there is power via the copper pair one can also feed with one or the other polarity and this can provide ad-ditional signals to the installation, e.g. a polarity reversal when the called party answers. These signals or short interruptions of the feed current are impossible to use with most electronic solutions.
An important point is that customers subscribing to one or another signalling possibility shall not be transferred to electronics which can not transmit the signal. In order to avoid this one must at first map which limitations the different electronic solutions will imply. We choose in the group to make a list of the basic services, such as polarity reversal, and whether they could be delivered by the different electronic solutions. We then made a list of customer services and their need for the basic services. The sales staff needed such a list when they discussed with the customers.
The work led to a report, the last one before Copenhagen Telephones was disbanded at the end of 1995. Within the Network Division Johnny Damgaard continued the work and added in 1997 another pair of solutions and their limitations to the report.
The Housing-group
In the middle of 1996 all delivery of telecommunication services in Denmark was liber-ated. The laws were not quite ready, this happened only a year later, but it was obvious that there would be several requirements to Tele Danmark as it was the dominating sup-plier. One was the requirement that others could demand to get room in the buildings of Tele Danmark for their equipment to obtain “interconnect”, viz. the coupling between the networks.
It would be a matter of consequence that there should be rules for such interconnections. The other operator of mobile telephony, Sonofon, already had this type of interconnec-tions, otherwise their subscribers could only connect to each other, not with the users of the fixed network. In that case there would be only a few people preferring Sonofon, even if calls in their network were free!
The starting point might have been similar interconnects as those Tele Danmark had with telecommunication operators abroad, i.e. connections via transmission systems between equipment at the premises of each operator and with agreed protocols for transmission and signalling on the connection. But as stated the requirement here went further, the operators should have a right to install their equipment in the buildings of Tele Danmark.
Different groups were formed to consider the consequences for Tele Danmark and pro-pose rules for “interconnect”. I participated in a group evaluating the consequences of the requirement for collocation and we met repeatedly during 1996.
Our starting point was that interconnect should take place at a rather high level in the network, just like the connections to operators abroad. A number of exchanges through-out Denmark were chosen in which other operators could have their equipment installed and be connected to the Tele Danmark network.
A problem was and is that Tele Danmark has obligations towards the Danish state con-cerning the safeguarding of telecommunications under all circumstances. For this reason it can not be allowed for everybody to enter the buildings. This caused another require-ment to the buildings where interconnect should be possible: They had to be divisible in an area with access for other operators and an area exclusively for the Tele Danmark staff. Collocation was therefore divided in three classes, either each operator had his own locked room with a common access from the outside, or the operators had a com-mon room with access from the outside, or – as the third possibility – the equipment was placed in the same room as the equipment of Tele Danmark. In that case it was required that there were always personnel from Tele Danmark present during access.
Luckily there were already examples of other operators of especially data services hav-ing their equipment collocated with Tele Danmark equipment. The agreements about this was a good starting point.
However, collocation could not be limited to those few places in Denmark. A new op-erator might be a supplier of TV distribution who extended his plant to replace the Tele Danmark access network, i.e. the network between exchange and the single user. Such an operator needed access at a subscriber exchange and the rules had to adapt to this. In fact it applied not only to the need for collocation (where access to a transmission sys-tem in most cases could cover the needs) but also the subscription price. Tele Danmark had to realise that the large expenses in the access network disappeared in this case, such that the new operator could charge the fixed pay for the subscription. But there was compensation: The new operator was responsible for the whole access network and Tele Danmark should not any more make up bills to all subscribers, the new operator was their sole representative.
In the end of 1996 the work was halted as there was doubt about the legislation and its consequences for Tele Danmark. Early in 1997 the chairman of the group was delegated to Newtelco in Switzerland, which had been partly bought by Tele Danmark, and he was later employed by Newtelco. Therefore the group was inactive up to the time I retired.
The MCID-group
I became part of another group in 1996 and 1997, but not as a proper group member. The group should make proposals for a change of the alarm service such that it would not be necessary to have a specially developed program for the AXE exchanges from LME to run it. MCID means “Malicious Call IDentification”.
The reason for not being a proper member was that this matter applied to exchanges and my department, NUI, had nothing to do with them, they were the field of NUS. But con-sidering my past I knew about exchanges and signalling, and therefore the group wanted my active participation.
The specially developed program saw to it that when there was an alarm call the AXE exchange switched in a tone receiver in parallel on the line during the whole call. When the alarm centre pushed a code on their telephone set the receiver caused a signal to the supervision centre of Tele Danmark, where all exchanges are supervised at all times, printing out the number of the caller. The staff here could find the address of the caller and – when the alarm centre called – tell it.
The standard program in AXE for MCID did not switch in a tone receiver constantly. It was only called in (and sent a dial tone) when the called party (i.c. the alarm centre) sent an R-key signal, a short interruption of the current in the subscriber line. The develop-ment of the special program for Denmark had a higher and higher cost for each new program package for AXE, so it would be an economical advantage to use the standard program.
But this was not only about an extra push of a button in the alarm centre, a push already introduced in several places in Denmark. Many alarm centres used a PABX, possibly connected via a 2 Mbit/s link to the public exchange. Thus, the real problem was whether the used PABXes could send the R-key signal to the public exchange? An R-key signal was already used in PABXes (from long ago) to call for a new dial tone to start a call-back and transfer. The PABX would thus have to, after having received an R-key signal, perceive a special dialled code as a message meaning that it should send an R-key signal to the public exchange. There would then come a dial tone from this exchange and the code for printing the calling number could be dialled. The user at the alarm centre should in this way both push the R-key and dial the special code in addition to what he was used to. This was not very ingenious!
In addition to this all alarm calls in Copenhagen go to the fire department, who transfers the call to the police if it is about other than fire or ambulance services. In this way the police might have to send three codes after the R-key signal before the last one gave a print-out of the calling number!
One of the solutions considered was to make all connections to alarm centres into ISDN connections. In this case there would at all times be a special signalling channel avail-able independent of the speech channel to send the necessary signal from a function key during the whole call. However, such a change was not cheap and might require new PABXes in the alarm centres.
Finally, the solution could not just be a proposal from the group introduced over night. It had to be field-tested before coming in actual use. This had to wait for the next pro-gram package for the AXE exchanges to test in connection with the standard program. Therefore the group had not finished its job when I – along with several others – left Tele Danmark.
Retirement
In the summer of 1997 the message came from the CEO of Tele Danmark: All employ-ees born in 1940 or earlier could terminate their employment at the same conditions as if they had been fired. This suited me fine, I would shortly be 65 years of age, the normal pension time if I had stayed with Philips, and the economy was OK. Further, like all oldies I could expect that if I did not resign, I would be fired! Thus, by the end of Au-gust 1997 my professional career ended...