>One of the few differences.That's your opinion.I don't think anyone
>knows yet because they are just starting with the
>standardization work. One main difference is that W-CDMA is
>interfacing a GSM network
Today a significant majority of the network interface is other
than GSM-MAP, with the IS-41 being the largest chunk of the
non-GSM interfaces. That is why QCOM fairness criteria explicitly
asks for standardization to support GSM as well as IS-41 network
interfaces to be supported equally.
Besides, with the GSM-CDMA trials, QCOM has already demonstrated
that it is possible to use a CDMA air interface with GSM network
interface. W-CDMA is designed for GSM-MAP (by the way that is an
issue in Japan), and CDMA2000 for IS-41. If convergence were to
occur, the converged common air interface could be made to interface
to both network interface.
>You say "the other 99% of the binders are largely redundant with
>cdma2000" I will not go in to details but I can give you one example
>why a 3g system is not just a broader pipe that effects the air
>interface. Today the signalling systems in the public network only
>support 64kbit/s for circuit switched traffic
Agree that there are significant differences elsewhere, but mostly
the higher layers are handled in SW and can be accomodated
more easily (synchronized base-stations and variable rate coders
are two big hw/sw exception to the above).
As to the more specific point about circuit switched data, I am not
sure what constraint you are refering to (ISDN is able to handle
more). In any event, why would you want circuit switch data above
64kbps? Going forward it clear as daylight that packet switched
network is the way to go, and both 3G CDMA standards promote this
model.
>This is why a converged standard will be so difficult to implement
>even if they could agree on converging the airinterface.To have a
>network that is both backwardcompatible to IS-41 and to GSM will be
>a huge task to achieve. QCOM is saying that they demands just
>that.It's easy for them to say.Ericsson is probably the company that
>has done most work on the GSM and IS-41...
It stands to reason that Ericcson has done more work on GSM than
QCOM. IS-41, may be also, I don't know. But the point is that on
the base station side you don't have to support both. The mobile
can roam into both GSM and IS-41 network interface areas. This is
not such an outlandish concept. QCOM has done field trials with
GSM-CDMA already. In any event what are the alternatives if you
want a phone to be operable widely/globally? Satellites only?
Multimode phones have to overcome the same problems. Think of the
option: you buy a phone for $x, and it only supports the protocol
supported in your home country. You want to go abroad or whatever,
you download the software over the air, pay some nominal amount
and you are good to go (to reduce charges you might only want
global roaming limited to only the duration of your vacation).
But I am not sure what is the larger point you are making. Is it
that convergence is technically impossible? Or that it is simply
undesirable? I think it is technically feasible (yes, standardization
will be nightmare, but in my opinion extremely valuable). Desirablity
of course is strongly dependent on your perspective. I would say
wherever there has been only one standard (music CD) in the consumer
domain, the cost has come down faster. |
