LTE vs WiMax: spectrum

[ilkhan]

In the US, each company has a nationwide network in pretty much one spectrum band. Correct me if Im wrong, but Sprint/Clear's WiMax network is all 2.5Ghz, while ATT/Verizon each have nationwide (apportioned variably by area) of the 700Mhz band and both are going to be deploying LTE. I have no idea what T-mobile's plans are, AFAIK they're doing HSPA+ and then jumping to LTE+ or something.

How are those choices (2.5Ghz vs 700Mhz) going to compare for range/build penetration? What else do we know?

 

[pm]

From a purely electromagnetic perspective, 700MHz is former TV spectrum and it's going to be great overall - excellent building penetration, excellent long distance over ground, even in hills and mountains. 2.5GHz is not going be as good - less range, and much worse building penetration because building materials become more "opaque" to radio signals above about 2GHz-ish (IIRC) and it's a more noisy chunk of spectrum (microwave ovens in particular) overall so more interference. About the only upside of 2.5GHz is that it's cheaper and less desireable and thus carriers pay a lot less for it and so it should be cheaper for consumers as well, and you can transmit more data over a given set of frequencies so you get more data bandwidth for a given frequency block.

 

[sswingle]

If you consider analog TV could be received 50+ miles away, thats damn good for one tower. Theoretically they can put LTE on fewer towers than 3G but get more coverage.

 

[lokiju]

WiMax with it's 2.5ghz already has proven to be terrible at penetrating buildings.

That very reason is enough for me to thing LTE is the one that'll be around when all is said and done.

 

[jersiq]

If you consider analog TV could be received 50+ miles away, thats damn good for one tower. Theoretically they can put LTE on fewer towers than 3G but get more coverage.

No you cannot. LTE (actually OFDM in general) is far more susceptible to multipath than earlier technologies. You can't hope to turn something up at that power and not expect multipath.

Besides, equipment vendors won't make amplifiers as strong as those seen for other terrestrial signals. No one would buy them.

Lastly, your uplink to the site would also have to travel the same distance, and I don't see any plans for portable amplifiers yet.

 

[Suspicious-Teach8788]

WiMax with it's 2.5ghz already has proven to be terrible at penetrating buildings.

That very reason is enough for me to thing LTE is the one that'll be around when all is said and done.

Right. Think T-mobile when it used only 1900 GSM. Terrible building penetration. You could take an AT&T and T-Mobile SIM and compare them on two phones indoors. It would be 1 bar vs 4 bars. This was also when AT&T/T-Mo had that fiasco in CA where they shared towers and then the Cingular transaction went through so they were riding the same towers for a good year. The difference between 850 and 1900 was phenomenal. I remember calling my friend as he walked into a mall in SF and his call dropped with T-Mo. LOL. I laughed as he had to stand outside to finish talking to me.

Anyway, the question is whether Canada and the rest of South America will pursue AWS frequencies too. If so, this would make buying unlocked phones a LOT easier. To mush up AT&T and Verizon into the same spectrum would give phone manufacturers a larger incentive to deploy NAM devices. Right now to get an AT&T compatible device, we're relying on Canada/South America/Australia. How much demand is that anyway? Not enough. VZW and AT&T have enough pull to get specialized devices, so I feel that if the entire Americas pursues AWS frequencies, we could be in good shape.

If not and they're sticking to 850/1900, I might as well shoot myself. Unlocked phones have been taking steps back ever since the deployment of 3G. I have yet to see many world-wide 3G phones (only iPhone, Nokia N8, some NAM devices aimed for Canada), but quad band 2G was already widespread by 2004. If we're gonna split AWS and 850/1900, then I'm guessing our chances of getting unlocked phones that work are like the chances of any of those phones announced on a daily basis will work with VZW currently (since virtually very amazing device ends up being GSM anyway).

It is already hard to get your hands on a 850/1900 3G phone, but to now split this market further into a 700/1700 world will just make manufacturers go crazy.

 

[JS80]

Where will VZ get the money to deploy LTE? They only have $3b in the bank with $61b in debt. Vaporware is vaporware.

 

[pm]

No you cannot. LTE (actually OFDM in general) is far more susceptible to multipath than earlier technologies. You can't hope to turn something up at that power and not expect multipath.

I thought the whole point of choosing orthogonal frequency division multiplexing as the signal bearer for LTE was it's resistance to multipath interference.