The author stated that he was very impressed with latency.

I don’t want to step on any toes of this site so I won’t post the link but you can easily find it by googling “sprint” then clicking on news.

Thanks for your response. A few additional facts:

1. With cell sizes of 1KM, this involved LOTS of physical buildout just to cover a single DMA, and with it, lots of hassles on getting antenna mounting locations that work – that is why it reminds me so much of mesh deployments.

2. The spectrum is not great spectrum – really poor building penetration (which we both agree on), and much more susceptible to ground effect signal degradation – basic RF engineering math supports this POV.

3. The average full deployment and adoption stats stand as quotes – targets of 2MB down/1MB up end-to-end.

4. There is no really material bandwidth spread from 3G that creates a material new consumer experience (I view the WiMaz deployments as patchy mid-band experiences – will be hard to move me off of that POV). I am not selling what LTE portends – this is a comparison with in-flight 3G deployments.

I think this should be a component of a larger mobile play for Sprint (i.e., incorporated in a hybrid data plan mixing 3G and WiMax for a great composite experience). Unfortunately, now that xohm is a standalone business run by Clearwire, this is much harder to pull off.

This is not a mesh topology. The backhaul is completely separate and I believe mostly wired.

You are quite wrong to be pessimistic about link speeds as compared to incumbent 3G. Xohm has 100MHz of spectrum available in their markets, unencumbered by legacy use. Even assuming LTE brings bits/HZ parity, Xohm will always have more bandwidth because they have more spectrum. Anyway, a real world speed of 2M down / 1 M up would be 3X faster than what you get with current 3G deployments (assuming at least a moderately loaded network). And 3X is exactly what I would expect, given that Xohm is deploying in 3X the spectral bandwidth that AT&T is using for HSPA.

Also 2.6GHz is fine for mobility. The real shortcoming of this network will be building penetration in urban areas. Given how poorly Sprint and the others have addressed this issue with their existing PCS networks, I’d say this is a significant concern. I think they will be counting on devices that can roam on to WiFi mitigate the need to deploy a lot of in-building WiMAX.

@Andreas

Latency should be far better than any US 3G build, because there is no interface with a legacy circuit switched core. The incumeants didn’t have to build their 3G networks that way, but they did because they like sticking with their comfort zones.

1-2MB download/.5-1M upload real throughput (stop quoting theoretical numbers – that’s just engineering propaganda) – don’t tell me in a full metro buildout you will get 4MB average download – the only way that happens is if you have no adoption (the constraints of the backhaul networks alone will not make this math work).

Consumer pricing that needs to be (post this initial promo) in the $40ish/month range (may be off $5 here either direction) to fund the buildout

All the complexities of mesh network deployment (1km-ish cells)

Spotty coverage even in the initial markets (will make T-Mobile look good ;)

Bad spectrum for a mobile play

Addtional CPE for home that a consumer has to pay for, one way or the other

No triple/quad play economics to afford the high CPA expenses

Stop me if you’ve had enough!

i think many people will be surprised not only at the growth of this network; but also at the speed in which users begin to drop their wired DSL/Cable connections.

also the big bulk of usage will not be on cool mobile devices like the nokia n810 but regular old laptop computers.