You know, their claims in that article are sooooooo misleading and contradictory. They claim 2 different data rates and frequency plans in the same article. Nowhere do they mention the all-important spreading factor they're using, but it has to be fairly large in order to accommodate a large number of simultaneous users, which means the "real" data rate is much, much smaller than the claims. Spread spectrum in general is extremely bandwidth INEFFICIENT. That's not where it excels. Loosely speaking, bandwidth is traded off for noise immunity. Since they don't give any meaningful data, we can only make assumptions. Since they say 5 channels, assume they limit subscribers to about 100 per frequency slot, or 500 max simultaneous in the upstream. As a conservative guess, use a spread factor of 256 over 100 orthogonal users per 6 MHz slot, giving raw rate of, call it 25 kbps. Gives 100kbps w/ 16 QAM riding underneath. And that doesn't account for overhead. Yes I'm guessing what they're doing, but I don't think I'm off by an order of magnitude. To me, it's very believable that COM21 achieves (and claims)greater usable bandwidth that Terayon simply because Direct Sequence Spreading is so wasteful of bandwidth.
Dunno what to think. Wish I had some real information instead of marketing fluff. They have a great marketecture--but what's the architecture. I'm beginning to believe that Terayon's sCDMA is misapplied in the cable upstream environment. I must be looking at something wrong because I can't believe they would think they could find widespread acceptance if my numbers are even relatively close.
dh |
