SUMMARY REPORT OF MEETING OF CCITT STUDY GROUP XVII GENEVA, SWITZERLAND, 13-21 MARCH 1989 by Toby Nixon, Principal Engineer and Standards Committee Representative, Hayes Microcomputer Products, Inc. INTRODUCTION CCITT Study Group XVII, which handles standards related to data transmission on the telephone network, met in Geneva, Switzerland, March 13-21, 1989. This is a summary of activity on items of widespread interest, prepared by Toby Nixon, representative to the CCITT from Hayes Microcomputer Products, Inc., and a member of the USA delegation at the meeting. V.42bis DATA COMPRESSION The draft recommendation for data compression in V.42 modems reached an advanced stage of development due to extensive and lengthy work by the editor (Mr. Alan Clark of the UK) and the ad hoc editing group. It was agreed to apply the accelerated approval procedures at the September 25-29 meeting of Study Group XVII. Under the new CCITT rules, this will result in a final, accepted recommendation (rather than merely a "provisional" recommendation) shortly after the meeting. Between now and then, manufacturers will be producing and testing prototypes to verify the correctness of the draft standard. INTERWORKING BETWEEN V.32 AND LOWER SPEED MODEMS The proposed annex to V.32 specifying interworking with V.22bis modems also received considerable attention, and a compromise was reached between the competing positions that appears to solve all of the outstanding issues. Again, between now and the fall meeting, manufacturers will be producing prototypes to verify the draft, and testing them with other multimode modems and with existing lower-speed and V.32-only modems. HIGHER SPEEDS IN V.32 British Telecomm presented very complete and persuasive test results that indicate the feasibility of 12,000 bit/s and 14,400 bit/s operation of echo-cancelling modems (V.32) over domestic and international circuits. However, it does not appear that their proposal is sufficiently detailed to permit adoption of an extended V.32 recommendation incorporating their techniques before at least the Spring 1990 meeting of the study group. Manufacturers will be examining the BT contributions and trying to reproduce their test results. The study group agreed that a half-duplex fast-turnaround operating mode for V.32 (similar to that in the Hayes V-series Smartmodem 9600) would be considered if the asymmetrical modem work was discontinued. A liaison statement was sent to ISO requesting their views on ways of expanding the V.24 (RS-232) speed indication/selection circuits to accommodate modems with more than two speeds. CORRECTIONS TO V.42 AND STATISTICAL MULTIPLEXING The proposed corrections to V.42 from the USA and Hayes and the V.42 statistical multiplexing extensions proposed by AT&T and Hayes were accepted in principle and text drafted, but time constraints precluded adoption of the final text. The information will be annexed to the report of the meeting, and therefore available to implementors and for comment at the next meeting (the corrections are already incorporated in Hayes V.42 modems). AT&T committed to perform a simulation of the performance of a V.42-based statistical multiplexer versus an existing statistical multiplexing design which packs data from several ports into a single information frame. ASYMMETRICAL MODEM The draft document which was jointly edited by USRobotics and Telebit was reviewed once again, and additional editorial changes made. The work is stalled awaiting test results. Unless comparative testing shows that the asymmetrical modes provide significant performance or cost/benefit advantages over V.32 (including the new higher-speed V.32 proposals), the work on the standard will be discontinued. A number of concerns continue to be raised regarding the effect of forward-channel far-end echoes on the backward-channel transmission in digital carrier systems (introduction of extreme quantization noise), and the use of frequencies outside the normal voice band (above 3100 Hz). INTERNATIONAL STANDARD "AT" COMMAND SET, V.25bis ENHANCEMENTS Toby Nixon was appointed as Special Rapporteur for Question 14/XVII, which is concerned with enhancements to Recommendation V.25bis (the current international standard for programmatic control of modems by computers and terminals). Based on statements from the chairmen of Study Group XVII, Mr. Klaus Kern of West Germany, and of Working Party 1, Mr. Richard Brandt of AT&T, the clear intention of asking Mr. Nixon to serve in this position was to get an AT command set-based CCITT Recommendation as quickly as possible. Mr. Kern, in fact, said that he strongly hoped that a version of the AT command set standard being worked on in the USA could be presented as a normal (white) contribution in advance of the September meeting of the study group. NEED FOR STANDARD "AUTOMATIC SPEED BUFFERING" MODE The need was identified for development of enhancements to V.14 (async to sync conversion, without error-control) to incorporate buffering and flow control. This is, of course, actually already done by most manufacturers of error-control modems, but is not provided for in any recommendation. Hayes will make a contribution on how this is done, including Transparent XON/XOFF flow control (already a feature of Hayes V-series modems). There are several applications for Transparent XON/XOFF and recommendations that could be affected: V.14, V.42 (error control), V.25bis (configuration), V.120 (ISDN terminal adapters), X.3 (X.25 PADs), etc. Also, V.14 and enhanced V.14 (with flow control and buffering) in non-error-control half-duplex and asymmetrical modems will be studied. FORWARD ERROR CONTROL FOR CELLULAR TELEPHONE MODEMS OKI of Japan proposed extensions to V.42 to include redundant encoding of various types to improve the performance on cellular telephone channels, which have very high error rates. The proposals are complex, and little work was done beyond the introduction of the documents. Delegates will study them, and more discussion will take place at the next meeting. NETWORK MANAGEMENT The activities which took place at previous Special Rapporteur's meetings was reviewed. The topic which received the most discussion was the current verbose encoding rules planned for management protocols. Study Group XVII sent liaisons to other groups regarding development of more compact encodings to allow efficient operation on the very-low-speed side channels used for network management communications between modems. The group also reviewed contributions proposing object-oriented descriptions of modems and terminal adapters, a necessary stepping-stone toward a physical-layer network management standard. ISDN TERMINAL ADAPTERS A contribution from the USA on changes to Recommendation V.120 to accommodate additional packet mode bearer services was reviewed and largely accepted. Further work was deferred until the next meeting because of the time devoted to other projects.