A History of JTG 4-9-11
A personal view, by John Pahl Joint Task Group 4-9-11 was born out of the arguments and deal-making at WRC 97. In the basement of the Geneva CICG conference centre I was part of the discussions that went on evenings and weekends for almost 4 weeks to shape the satellite systems of the future. The driving force was the boom in satellite communications: after the voice and narrow-band data of MSS systems the next stage was for high data rates, broadband services - the so called "Internet in the Sky". There was excitement at the idea of using non-GSO rather than GSO orbits to provide global coverage and get shorter delay times. However the most useful frequency bands were also heavily used by GSO systems, and the Radio Regulations said that non-GSO systems would have to protect the GSO and not cause "unacceptable interference". This phrase was considered problematic as "acceptable" levels were not defined but would have to be agreed during negotiations with the many parties involved. The job of WRC is to revise the Radio Regulations, including clearing up problems such as this; the debates are about how. Europe and the US came to WRC 97 with two very different visions about how to proceed. The US wanted to build on the previous conference, WRC 95, where 400 MHz in Ka band had been allocated to be available for both non-GSO and GSO systems on the basis that the first to apply for use, to file, gets priority over the next. The first to file was the US non-GSO system, Teledesic, and they wanted to increase the allocation to 500 MHz and make it retrospective to WRC 95. On the other hand Europe proposed to open up the lower Ku band to non-GSO system by quantifying what "unacceptable interference" meant by specifying limits in terms of the EPFD, the Equivalent Power Flux Density. Any non-GSO system that produced less than this interference would have met its obligations under the Radio Regulations. Neither were happy with the others view. Europe didn't like allocating more spectrum to non-GSO systems in a way which would make it very hard for GSO systems to operate in the same band. The US objected that the values of the hard limits proposed by Europe hadn't been properly studied and that GSO use of the Ku band was too important to take any rash steps. The stage was set for a classic WRC battle. After many formal debates in committees, corridor discussions, secret and not-so secret steering groups, a deal was hammered out for both sides to get what they want. The US got its additional 100 MHz in Ka band, while Europe got its hard limits in Ku band, though on a provisional basis, with the ITU-R studying the values themselves before the next Conference. The forum for this would be a new group, a Joint Task Group, the JTG. It would look at how to protect a range of services from the non-GSO FSS, namely the GSO Fixed Satellite Service (the remit of ITU-R Study Group 4), Terrestrial Fixed Service (Study Group 9), Broadcast Satellite Service (Study Group 11), and others including Space Science and Radiolocation. From the numbers of the relevant Study Groups came its full title: Joint Task Group 4-9-11. JTG Meetings So the JTG was started, and a chairman agreed, John Leary - a fellow Brit. working for Japan's SCC. Soon we had terms of reference, web sites, email distribution lists, and papers - lots and lots of papers. In total there were:
The JTG formally opened its first meeting on 9th March 1998 in Geneva. Following meetings were a great circuit of Toulouse (July 98), Long Beach (January 1999), before finally finishing back in Geneva in May 1999. There were other ITU meetings too - WP 4A, WP 4-9S and JWP 10-11S in between these which contributed much to the work done. Also in this small world there were many informal meetings, where the main players could negotiate and discuss behind closed doors. At each meeting there were many familiar faces that I saw more often than some friends back in London. It wasn't all work: in Toulouse we were entertained by the French delegation to dinner at a beautiful Chateau, with entertainers and fireworks. Later that week I was in a square by the river in the old part of town as France reached the finals of the World Cup, surrounded by thousands of students from the local universities. It was an amazing scene, the noise, the excitement, flame throwers and musicians, and everywhere cars honking their horns. In Long Beach we were invited by the US GSO coalition to the Queen Mary moored across the bay from the conference centre. First there was a tour of the ship (in which we managed to get lost in our hunt for the engine room), then dinner and dancing to a band. At the last meeting in Geneva we celebrated my birthday by dinner with the UK delegation at a jazz restaurant called the Halle d'Ille, followed by clubbing with the US delegates. I also managed to get some personal travel time, going from the first Geneva meeting to Turin, and using a few days before Long Beach to explore LA and the nearby Joshua Tree National Park. But during the meetings it was very hard work indeed. At Long Beach I worked 13 hours a day for almost all of the 10 meeting days. Often drafting groups and technical discussions went on late at night - not just for one day, but day after day. The final session of WP 4A ended at 5 in the morning. Down to Business There were so many questions to look at, so many different ideas to try, assumptions to check, methodologies to develop. We built up a new set of acronyms, code words for our little world, where we knew why the limits were for EPFDdown/up/is, the difference between Methodologies A to D, what was 1323, 1325, 1328, in-jokes (such as Francois Rancy, head of the French delegation, correcting the English of the UK delegation). There were changes as people moved on, changed jobs, got married, had children, retired. Representatives from 20 to 30 countries plus international organisations had a job to do, a task that could effect business and communities all over the world. Some of the questions were looked at included:
There was too much to follow each in finest detail - we each had our own set of pet subjects of interest, and left the rest to others. The Software Specification My special interest at the JTG was related to developing a software specification. EPFD is not just interference at a particular moment, it is a distribution, a range of interferences associated with how often those levels are exceeded. This is needed as satellites in non-GSO constellations move with respect to Earth and so give interference levels that vary in time. So calculating the EPFD that a particular non-GSO system would generate is not simple, but requires such extensive calculation that software must be written. The first problem was to decide the basic approach used. Initially Tony Reed (Chairman of WP 4A) and I had been thinking about "Full Simulation", whereby each beam of each satellite was modelled in detail, including factors such as tracking strategies, gain patterns, loading, and earth station locations. However operators of non-GSO systems didn't want to describe some of the commercially sensitive aspects to their system, and also wanted the flexibility to change their systems slightly without having to re-apply to the ITU. Instead they proposed a "PFD Engine", where a program, a "black box", would generate PFD for each satellite for each time step for a specific GSO ES location. There were many debates about this at the first Geneva meeting and again in Toulouse. One problem with this approach was that there wouldn't be full disclosure of the non-GSO system so that EPFD levels could be independently verified. At Toulouse I gave a presentation showing how our software, Visualyse, could be used in either method. However even as I did so, a third option was being discussed, to use worst case assumptions to generate a PFD mask on the satellite and then use that in the simulations. The "PFD Mask" approach meant that all the key parameters of the non-GSO system would have to be available to generate the mask itself, but as some worst case assumptions were made there would be no need to reveal some specific aspects such as detailed resource algorithms. Additionally minor operational changes to the system could be made providing that they did not result in the levels in the PFD mask being exceeded. Finally the PFD generated by each non-GSO satellite could, in theory, be measured operationally. So this compromise was accepted as the way forward. However writing a detailed specification for such software was clearly a task that required more time than was available just at these meetings. So John Leary set up the Software Group of the JTG, and asked me to be a member, the contact point for the UK. Also on the group were a representative from the US, France, Canada, Japan, Russia, and the ITU-R/BR itself, plus Brazil (as Dr Fortes chaired the group at JTG meetings). Over the next 6 months this small Software Group worked by email, gradually putting together the various parts of the Software Specification. I was responsible for one of the most important parts, Section D, the algorithm to calculate EPFD itself. At Long Beach the document had to be integrated, presented to the full JTG, revised with comments, and finally approved. As well as being involved in discussions, revising the document, developing key algorithms, I was also chairing one of the two sub-groups on this issue. It was a very busy time! But the work was not quite finished. At the final meeting we were still revising, making final changes, filling in the gaps. It had grown to be a monster, 105 page, document. Together with the others of the Software Group, and in particular Dr Fortes and Marie-Noelle Thouvenin from France, we nursed it through the JTG, then JWP 10-11S, to be eventually approved at SG 11. It did feel like a weight was lifted when it was finally approved. But in the "real world" outside the ITU, we still had to finish implementing the full algorithm in Visualyse, and now there's testing, and review by the BR. The Debate Concludes The JTG's most important question was "What power limits should there be to protect the GSO FSS / BSS and the FS?" To the last hour of the last day the JTG debated the hardest of these, the EPFDdown limits for the GSO FSS. There were two views: the gap between them had been reduced to a few dB but not closed. The room was tense as almost 18 months of work built up to a final debate. But it turned out we weren't able to walk away then with all issues closed. However the list of agreements is long, too long for this document. Those who are interested can read the Chairman's reports and the text to go in the Conference Preparatory meeting (CPM) report - they can be found on the ITU web site. A massive amount of work was done and the JTG achieved much. There are those that say there were too many meetings, not enough time for study and analysis, to review the work done by others. That may be true, especially as it seems hard that there should be a whole year between the end of the work of the JTG and the next WRC. But one of the keys to the whole process is getting people together: the corridor discussions, unofficial hints about possible areas of softening, explaining complex issues over a white board, and sharing data from laptops. Postscript - Before and After So we finished the work and went our separate ways - for a time. Since the JTG closed we have had CPM 99, the first WP 4A of the new millennium, and finally WRC 2000. At CPM the EPFD limits were finally agreed and so 4A had a lighter mood, a feeling of back to normal after a storm. The software spec. was still under discussion - as we implemented its algorithm to build Visualyse/EPFD we inevitably found minor gaps, ambiguities, and a few errors. This wasn't the first time - in the autumn there had been other minor changes at JWP 10-11S. At the Radio Assembly just before WRC 2000 the corrected version, the last document of the meeting, finally became a Recommendation. I was the only member of the Software Group there, to see its final acceptance, to know it was done. The story of the development of the ITU regulations for non-GSO has no neat start and end, though WRCs can be used as chapter start and ends. Of course non-GSO FSS didn't start at WRC 97 or even WRC 95. But its hard to give one time, one date when there are so many characters, roles, and places in this drama from the world of the ITU-R. I remember in 1996, at my first ITU meeting - WP 4A in Rio de Janeiro - seeing simulation results for a system called F-SAT MULTI 1B, the ITU name for SkyBridge, using arc avoidance to protect the GSO. When I look back at those papers - basic simulations showing distributions of C/I - I can feel how far we have come, how much work has been done. In May we went to Istanbul for WRC 2000. Here under blue skies and on airy terraces, in smart new conference rooms, at parties in palaces beside the Golden Horn, and on boats floating under the Bosphorus Bridge we met for the final meetings. The pressure was off and we could mostly relax - indeed I left early after 2 weeks. Where-as at WRC 97 the non-GSO battle had dominated, here it was the dog that didn't bark - and we were all relieved that all that hard work had been accepted. The next five years will be different from those between WRCs 1995 and 2000, as non-GSO proponents become more focused on providing and marketing a service. Then we shall see if it was all worth while. |