Frank,
Mr. Pluvia just forwarded me your TERN post to which this is a reply, this morning. Having posted TERNS latest news, and having seen the new products, suddenly I realized when reading your post that I might have something to offer. I hope it isn't a complete bust, as you know, I've been struggling here. But I'm curious, no doubt. I would like to offer a bit of information that hasn't been discussed so far as I know, and bounce my suspicions off the wall as it were; here- where the threads of fiber are a little quieter. This post asks the kinds of questions I wanted to ask on the CMTO thread- but I wasn't as prepared to figure out how to do so as I thought.
Re: "IMO, S-CDMA over cable can only be regarded as a means of
prolonging the agony of analog for those subscribers who reside in
the territories where cable operators use them, over a much longer period
of time than a natural death would suggest. Where TERN's modems (and
many other cable vendors' wares are concerned, but we're focusing on
TERN here) are installed, users will have to live with black coax, without
the benefit of fiber, in a bandwidth restricted mode, ad infinitum. That has
been a selling point of theirs, btw."
As I've suspected, TERN is not standing still. I've provided a link in the past on the TERN thead in which some folks have indicated S-CDMA over HFC is a good idea. Let's just say that eventually, my nose led me to question whether S-CDMA is as tame as suspected here. Last I began to address this(in an unwelcome and heated defensive state of mind over Pat Mudge's displeasure with me and my own perhaps virulent replies- which I regret), responses I received from Mark Laubach seemed wanting- as he'd finally allowed that S-CDMA had an advantage that "may" not be worth it, despite prior statements which I'd felt indicated there could be no S-CDMA advantage no matter what. This left me in a state of greater doubt still, and other points I offered went un-commented upon which I felt might be significant-- but I digress beyond my understanding.
As you may know, Mr Gilder has said, essentially, that he feels TERN coupled with fiber to the curb and coax to the home, ought to be compelling for about 10 years. After that, he seems to have grave doubts about Cable modems period...but thinks a saving grace could be that Cable Co's might likely be the first to deploy FTTH. Having said that, you may ask, why would HFC and TERN go together as Mr. Gilder supposes?
TERN has apparently recently begun offering new IP telephony and broadband local wireless, etc., requiring and intended for HFC. This would seem to mitigate the notion that Gilder, by "...promote(ing) TERN in this context is to effectively justify the perpetuation
of a technology which preserves the embedded investments in analog
technologies of the cable cartel, which will (subtlely or otherwise) create a
huge roadblock to the eventual fiberization of those sections of America
and the World who use them."
The bandwidth of TERNS newest products intended for HFC is limited, so maybe I guess I'm not really addressing your concerns if you mean to insist on a need replace coax altogether with fiber, but I know you don't expect that. So when you say "users will have to live with black coax, without
the benefit of fiber," I'm immediately thinking, in terms of TERN, and realizing there are significant if still slower? services that are now offered with HFC using S-CDMA, allowing TERN to offer low costs to get going for those who DON'T care to spend to go HFC for years to come, and decent services at low implementation costs for those who do.
I guess the question is, would we be satisfied for now with TERN and the others all driving HFC? Given the surprising success(to maybe several here, at least) of TERN, so too I can imagine the newest S-CDMA HFC solutions becomming quite successful- and at least, likely not delaying the extension of Fiber where everywhere S-CDMA is used, as you otherwise reasonably considered.
To answer you question as to what's more legacy than cable infrastructure, that would be the Gildered copper cage(pardons, ahem), of course. But you devil...Barring Sattelite and wireless solutions...we WILL have to live with these things for awhile, NO? Acchhh...Disruptions I can't imagine? Sooner Rather than later? I hope you are right- I'm sure we all do. If TERN is offering decent IP telephony, wireless video-conferencing, and a GSM phone that works with the modem when your are near home, etc. will they compete awhile with disruptors?
terayon.com
>>Terayon 's Multigate telephony and data access platforms deliver efficient, carrier-class voice and data services simultaneously over cable, based on robust, high-capacity S-CDMA technology. Multigate systems support a wide range of broadband services from hich operators can choose four levels of capability:
a base configuration with telephony (two lines)and cable television;
the addition of high-speed Internet access;
the addition of cordless voice and data based on the digital European cordless telecommunications (DECT)standard;and
the addition of video conferencing services.
Multigate delivers all these services from a single integrated headend box over one video channel to a single integrated subscriber unit, saving equipment costs and space. Multigate is the leader in fixed-mobile convergence (FMC)over cable, combining the cordless mobility of DECT with wireless data and telephony services for in-home networking. And when the DECT phone travels out of range, it automatically switches to GSM and becomes a cellular phone. Multigate features include centralized server management with advanced reporting and billing capabilities,support for multiple QoS levels,support for additional lines, and the flexibility to accommodate future technologies and provide a secure path to an end-to-end voice-over-IP solution, protecting your investment.
Multigate is an innovative local access system that allows emerging telecom service providers to deliver simultaneous voice and data services over cable TV infrastructure. Multigate supports telephony, high speed data (Internet), cordless voice and data (DECT), fixed-mobile convergence (DECT-GSM), ISDN, videoconferencing and other broadband services, all over one video channel and via a single integrated subscriber unit. Multigate is a highly cost-effective solution for residential and business markets.<<
Re: "Qualcom, as you both/all know,
negotiates the near-infinite realm of free space, which has a near-infinite
level of head room. TERN, on the other hand, deals with the limitations of
coaxial cable plant. These are two entirely different domains of conquest."
Yes, I've read that here before, and while the issues may be beyond me, I've bolded points below to highlight issues relating to the use of sychronus CDMA in CABLE VS plain CDMA in the air. I wonder if the following has thoroughly been discussed and/or discounted.
de.infowin.org
"An access system with a shared medium, such as a hybrid fiber coax (HFC) cluster or a radio cell, needs a multiple access scheme in the upstream direction to manage the traffic of a number of simultaneous users. ...Both frequency and time division multiplexing are mature telecommunication technologies, but are sensitive to ingress noise and interference. CDMA on the other hand is based on spread spectrum technology, which has been used in military communication for decades because of its resistance to jamming and eavesdropping, and has distinct advantages in media subject to ingress noise. In a coax system within the designated uplink band of 5-40 MHz, the ingress consists mainly of impulse noise, narrow band signals such as short wave and amateur radio. ... In addition, CDMA offers a certain degree of privacy through individual user individual codes and provides flexibility for system evolution as the demand for capacity and services grows. ..."A CDMA system has two possible operational modes. In asynchronous mode, the spreading sequences of different channels are arbitrary in phase. Here, the spreading gain is partly consumed by multiple access interference, since the codes are not exactly orthogonal, and the system capacity is limited. In synchronous mode, the spreading sequences of different channels have fixed mutual phase relations due to transmitter-receiver closed loop controls. Appropriate selection of codes allows the MAI to be minimised and offers a significant increase in system capacity.
CDMA in its classical form provides continuous transmission at the same fixed data rate for all users, whereas ATM data streams are bursty packet oriented signals with constant or variable bitrate.("bursty"--Gilderesque?Dan B) Based on a set of alternative flexibility schemes, FLEXIMACS will develop an S-CDMA functional specification and reference system configuration.
Appropriate solutions for MAI minimisation will be investigated. This will include the identification and specification of code families with correlation properties suitable for S-CDMA and advanced MAI cancellation strategies in the receiver. Physical layer control as well as a MAC protocol will be developed to map the ATM flexibility requirements onto the S-CDMA physical layer capabilities. They will be designed according to efficiency, delay and flexibility criteria taking existing TDMA standards into consideration... <<
On second look, this may not be getting at my curiousity here...and I can't find the right reference...Um...with CDMA in the air, signals in the air arrive at the receiver at different strengths- causing an adjustment problem. But with S-CDMA over coax, the S essentially causes all signals to arrive at the headend, um equally? eliminating that adjustment problem that a headend would otherwise have to deal with? I'm asking then, are we all aware that's not a problem wired S-CDMA has to deal with? THAT's it, from memory anyway, I hope. I'm afraid that was much ado about nothing, now that I think more of it. Well, I'm leaving it in the event someone learns something. I hope something in this post is redeeming, if not this question.
In the following I've put many questions in the bolded parenthesis.
internettelephony.com
>>Upstream advantages
For any given set of noise conditions, an S-CDMA channel provides a higher-reliability upstream channel than a TDMA channel. This is a result of S-CDMA's ability to handle dynamic narrowband interference and impulse noise, meaning that fewer interruptions in data transmission will result from excessive error rates. S-CDMA's ability to handle the same level of noise with fewer errors and channel outages increases service reliability for subscribers(Mark Laubach said Guarantee of service was S-CDMA's claim- but that it may not be worth it. Is it? He also told of types of noise which S-CDMA is particularly good at handling- I asked how much of typical plant noise is of those kinds, and didn't get a reply, but I recall- it's somewhere in a GTR- Gilder said typicaly up to 70 or 80% is of those types, if I remember the number closely, and I hope I do, it was big--so what might this realization mean in comparing S-CDMA to even highly filtered TDMA equipment, both over HFC where such noise may still exist(may it?), especially, I suppose, intermitently?- Dan B) , which in turn, can reduce operations and customer care expenses. In addition, advanced services, such as packet-based and circuit-mode telephony and videoconferencing, require a higher level of channel reliability than casual Internet browsing. ...
...Under any common channel reliability criteria, an S-CDMA channel provides higher payload capacity than a TDMA channel and serves more residential data subscribers per channel.(I recall Mark on CMTO having difficulty reconciling the reported no of homes S-CDMA deployments were reported to have passed by TCA, I believe. In reading Pat Mudges notes of CMTO's recent conference call, there was a statement to the effect that CDMA would require using an extra channel; implication-waste?. But if each channel indeed has more payload capacity, well, I guess you get the explanation of the homes passed...and equipment cost benefits as follows here?). This reduces capital expenditure for upstream receiver cards when building out the headend infrastructure and ensures efficient usage of headend equipment. The higher channel capacity also supports premium services(which I guess is behind TERN's ability to offer it's now S-CDMA systems?), providing guaranteed minimum and high peak upstream rates per subscriber that are substantially greater than those offered by ISDN or asymmetrical DSL. This, in turn, positions cable operators to obtain lucrative high-margin business selling telecommuting services and commercial-grade data access services.
The higher capacity channels in both current and developing cable modem specifications allow for much higher peak burst rates per subscriber, creating a unique competitive advantage for cable-based broadband access. <b?Higher upstream capacity channels will be necessary to provide voice and videoconferencing services at reasonable penetrations.(perhaps offsets Dave's argument that the 50/50 split of the upstream/downstream with S-CDMA is a waste simply decreasing subscriber speed?- and other arguments about TDMA handling bursty data just as well as S-CDMA?--Dan B)...
The amount of actual spectrum an operator needs to allocate to a given service also is important. The use of frequency hopping by TDMA systems to avoid noise interference requires an additional "clean" spectrum. For example, a single revenue-generating TDMA upstream channel with at least one additional channel width of spectrum lies dormant.(does this amount to some level of reversal of the concerns CMTO expressed about S-CDMA as I noted above?--Dan B) In addition, TDMA systems require that operators avoid known interferences such as CB and ham radio bands. Thus, the channel availability ratio--or the amount of usable spectrum relative to the total available--decreases rapidly with the width of a TDMA carrier. S-CDMA provides greater robustness in the face of such interference. ...<<
Well, I'm sorry for the long post again...But it took me awhile to get this out, obviously, and I'm genuinely curious, of course.
Oh...one more curiousity...and again, I can't find the right Ref. just now...but I do have this:
Capacity and tiered services
The S-CDMA Media Access Control Layer (MAC layer) minimizes intercode interference via a synchronized spread spectrum method, thus approaching the theoretical channel capacity limitation. The MAC layer?s cell-based architecture enables multiple tiers of service (guaranteed and best effort data rate) with the flexibility for symmetric and asymmetric bandwidth provisioning. This enables allocating bandwidth for each network modem dynamically, based on service provisioning, priority, overall channel capacity availability and real-time bandwidth use, so operators can offer Quality of Service (QoS) levels to its subscribers.
P.S. I haven't a clue why there was no press release concerning these new products. Perhaps they are not quite on the market yet? Maybe TERN's been reading Gilders recent freedom of information piece in the Wall Street Journal?
Freedom Works,
All the best,
Dan B |
