On Mon Dec 2 20:20:48 1996, Allan Angus , from None (former TR45 participant) , wrote:
In response to Tom Brush's comments of Dec. 2, the original 40x analog capacity claims for CDMA were made by the good folks at Qualcomm to CTIA and TIA in early 1989. I attended a presentation hosted by CTIA near the O'hare Airport in Chicago in May of that year, at which these claims were made, in connection with a truly amateur analysis of the nature of AMPS, TDMA, and CDMA capacities. In January of that year, TR45.3 had voted in favor of TDMA as the access method of choice (in fact I cast the first ballot). Since, at the time, CTIA were requiring TR45.3 to provide a method that would improve capacity by 10x, deliver the standard by 12/31/89, and not change the RF structure of the air inteface (all of which is in the CTIA User Performance Requirements document of 1988) changing horses in May of the year was not an option. It was certainly ill-supported by the 40x analysis that Qualcomm made in Chicago. TR45.3's stated position of the time was that a combination of second generation speech coding (to get 6x), improved re-use (N=7 -> N=3 or 4, to get another 7/3x), and trunking efficiencies were the methods of achieving 10x gains. In the short term, TR45.3 would deliver 3x, plus 7/3x, plus trunking gains. Hughes' packet voice (E-TDMA) was not on the table at the time, and did not play any role in the decision-making. In the intervening time, the TDMA cellular standards have been compelled to tes t to completely unrealistic delay spread requirement that the CDMA standards groups refuse to even recognize. This has been responsible for an unrealistic degradation in the performance of TDMA equipment on both sides of the air interface. The degradation is perceptible in, voice quality, in hand-set cost, and in the re-use numbers that TDMA was reduced to from the original plan. Ask a CDMA proponent how IS-95 equipment complies with the added delay spread requirements of the modified CTIA UPR, and they go somewhat quiet. This is especially interesting since it was service providers who are now firmly in the CDMA camp who demanded that TDMA equipment meet the extreme delay spread requirements, and who turn a blind eye to CDMA's performance on this issue. As to the whole matter of speech quality as a deciding issue between CDMA and TDMA, excuse me but these are access methods, not speech coders. If someone has a good speech coder, then it could be placed in either TDMA or CDMA equipment. The introduction of speech quality is a red herring in a discussion of access methods. There is much that has gone wrong in the TR45 standards committee since 1988 and in the cellular industry as a result, this I can agree with. I put it down to an utter loss of focus and nerve in getting ahead with the original job, which was abetted in large measure by the TDMA/CDMA debate. Frankly, it's largely a wash between FDMA, TDMA & CDMA. But we have spent millions of dollars fracturing domestic wireless R&D efforts over the last 7 years on a completely fruitless battle, while ignoring anything that meant performance to end-users: improved speech codecs, Q.931 dynamic call models, internet connectivity, serious support for circuit-switched data, etc, etc. This has all been such a total waste of effort. I do find it humorous though to see comments like Mr. Brush's, here in 1996, wondering where 40x came from. I guess you had to have been there.
|
