latest WiMAX status

A few years ago there was much whoopla about WiMAX being the best thing since sliced bread and physical affection between married couples.

I did not believe this. What is the present situation on 802.16? It looks like a slight improvement over 82.11 but I do not see terminals for it (either stand alone or built into computers) being sold.

WiMAX or 802.16e is a relative short path from research to deployment (5-8 years). The system themselve address many issues of the current systems.

It is

1. OFDM for some variability in rate/SNR
2. OFDMA terminal selectivity by frequency/time synchronisation instead of code
3. Frame based (basestation orientied, deterministic link, flexible rate, sleep frames)
4. MIMO built in

But more important is that the patents are spread over 10 times more companies than the competetive system. That lower the cost and timing of final system deployment.

Currently there are only small system deployments for 802.16d-2004 based equipment in some emerging countries.

It takes 3-5 years before every portable computer have a WiMAX link. And wireless DSL in non-cooper countries will be the entry followed by mobile DSL.

We just released a chipset for WiMAX and there are a few potential customers on the line.
Myself I am very skeptical in the full success of this system especially of the mobile one, 802.16e.
I mean full success, to work at specified throughput in a live network, otherwise there are many other proved systems working fine at lower speed.

The same enthusiasm was at the beginning of the first CDMA system (former IS-95) when everything looked fine in the lab but like a sh.t in a live network.

Every new system got some hype. That seems to me natural now after learning that the peaple which do not share the technical background only recognize the best case features of a system. Now com systems have so many feature dimensions that also for a experienced technical backgrounder it is difficult to answer how practical are these features are for the final user.

I think in comparison to code spread based MA system there are improvements. Taking for instance the MA separation within the same band OFDMA is in favor because separation in CDMA requires an increasing code length for more separation. In OFDMA it requires only a non-dataframe technical requirement of frequency synchronisation about the terminals. Because that is an other parameter dimension the logical data structure and rate granularity is not affected in OFDMA. Channel filtering is n*log2(n) versus n^2. MIMO is easier to implement and finally the data granularity is huge. It could be from a small fraction of the subcarriers for MA at BPSK up to all subcarriers at 64QAM. And that for each terminal individual.

The issues seems to me now that mobile channel estimatation with rolling pilot subcarriers is not fully tested in field. Also the big flexibility on the MA modulation is difficult to handle at the MAC layer. Beause the experience with the system is fresh improvements could be seen there.