[g.lite standards]
Hi Shlomi
seems that some of the nervousness out there might be related to questions about how orctf is positioned in regards to potential "new" adsl standards for g.lite (splitterless adsl)...
it wasn't clear at first if the intel/microsoft/compaq consortium (now uawg: universal adsl working group) was going to try to hijack the standards process or work within the existing system...
Also there are a number of competing schemes out there getting a lot of press, and it would not be good news if such an alternate scheme were to be adopted as "standard"... those co's like orctf with silicon in the pipeline would have the most to lose...
it also seemed for a while there that the the standards process could be compressed into a small window (once "announced" as universal adsl product by xmas 98). Such a situation would also seem to increase exposure for those who have already committed themselves to a certain course of engineering (could be back to the drawing board if standards evolved quickly in other directions)
only time will tell whether or not orctf will have problems with some of the details of the standards as they evolve, but it appears now that the direction and context of the process should support orctf's efforts todate... basically it looks like when the dust settles universal adsl will morph into g.lite which is going to be a subset of the existing standard (T1.413)... which means dmt line code... which is orctf's specialty and where they have placed their bets (next generation adsl silicon).. undoubtedly there will be lots of twists and turns along the way, but in general terms things seem to be headed in a positive direction...
also the timeframe for developing g.lite standards seems to have stretched out removing the risk of being blindsided near term.
the uawg will "develop a set of contributions that build on the existing American National Standards Institute T1.413 standard and propose it to the International Telecommunications Union's G.lite subcommittee" sometime in March(?)..
.. "An approved multicompany standard could be available before the end of 1998..."
techweb.com
"The group (uawg) says Universal ADSL modems will be available by 2000 ."
techweb.com
altho detail is scarce, after a quick first pass, it looks like the g.lite dmt silicon players furthest along are the same players as those for dmt adsl with one very important newcomer
dmt adsl silicon players : alcatel, adi/aware, fujitsu/orctf, pair, and ti...
adi/aware and fujitsu/orctf are the only parties to have publicly shown working splitterless prototypes...
pair and ti have announced silicon that will support g.lite...
alcatel has been pretty quiet... have only said that they "will work on implementing g.lite in its DSLAM (DSL access multiplexer) equipment."... but they undoubtedly have/are-working-on something.
and new to the party for g.lite: lucent... and they look real serious...
lucent: "... Lucent Technologies' microelectronics group threw its hat in the ring with a proposal for "WildWire," a splitterless ADSL with 1.5-Mbit/second downstream capability, based on discrete multitone (DMT) coding. Lucent is already showing prototypes of a complete chip set with general-purpose DSP, data converters and line drivers...
Lucent will bring several proposals before G.Lite "at the appropriate time," Brizel said. He added that chip sets will sample in the third quarter, and that many system products may arrive before a standard exists...."
techweb.cmp.com
"Lucent Technologies Inc. will tout its WildWire splitterless DSL chip set, which offers speeds of 1.5M bps and includes 56K-bps technology. The chip set will cost about $50 and is due in the third quarter, said sources close to the company..."
Message 3255861
and for context:
"...Bell Labs, Lucent's research and development arm, invented DSL technology and owns several fundamental patents..."
biz.yahoo.com
look at the arrival of lucent as both good and bad news... the bad news is that they are going to present very serious competition (for everyone)... the good news is that they will keep everyone honest in engineering terms, on track for dmt g.lite and on track for a (hopefully) nonproprietary solution... after the 56k wars lu knows only too well what is at stake... also they know orctf well and have interests aligned in other theatres... ..
One of the things that has been bothering me about the orctf/fujitsu splitterless press release is the upper end downstream speed quoted at 1Mbps... both lu and awre are using 1.5 Mbps... don't really think this is a big deal as am guessing (hoping) this is probably just a function of the engineering approach orctf is starting with (probably tones used etc) and maybe related to trade-offs for signal quality and distance...
if you get a chance to discuss this with orctf could you ask them:
1. Will the orctf solution be capable of 1.5Mbps? if so when?
2. If not, why they are not using an approach capable of such speeds?... after all believe they know this stuff forwards and backwards, and politics and posturing aside, g.lite seems basically to just be about "filtering and tweaking" signal ... in any case, believe they will need to be able to at least publish comparable numbers to hold their own in the public relations (disinformation) wars ...
... also if it seems appropriate would you ask/beg for an:
3. update on adsl silicon due mid 98? on track?
... i will try (again) via email, but recently seems to be hard to get a response...
thanks,
s |
