Today’s 4G World

LTE is solid, it works, it provides the best data speeds yet, and speeds will be getting better in future releases of the LTE specification. However, it is also prudent to understand that there will be times when a customer will fall back to 3G data levels.

Clearwire wants to switch to LTE and added 1.9 million wholesale subscribers this past quarter, Sprint is to roll out LTE on its 1900-MHz spectrum, AT&T lit up LTE in a few markets with more to follow, T-Mobile’s 4G system continues to expand, Verizon continues its mad dash to the LTE finish line, and LightSquared continues to try to bully the GPS industry and the FCC.

Clearwire has publically stated that it wants to deploy TD-LTE (Time Division LTE), which is a flavor of LTE that uses the same portion of the spectrum for both cell site and device transmissions as its WiMAX system does today. TD-LTE is being pushed in China as a standard, Qualcomm’s chipsets already support both FDD and TDD LTE, and those making WiMAX devices for Clearwire could easily convert their products to TD-LTE.

Clearwire currently has a lot of spectrum assets. On average, for all of the markets it has licensed it has 150 MHz of spectrum in the 2.5-GHZ band. This compares more than favorably with both AT&T and Verizon that hold licenses that average 91 MHz for each market. This spectrum is not as valuable as spectrum at 700, 800, or even 1900 MHz because building a network at 2.5 GHz (2500 MHz) requires more cell sites per area covered. On the other hand, because there are more cell sites, the network has more capacity available.

The issues for Clearwire are the cost of the transition and the timing. First, the technology needs to reach a point where it can be deployed. Next, Clearwire has to be able to afford to make the transition, which means it needs to raise more money. It could sell some of its spectrum and has stated that it is in talks with other network operators, although Verizon denies it is talking to Clearwire. However, TDD spectrum at 2.5 GHz is not worth as much as FDD spectrum at lower frequencies. If Congress passes the spectrum auction part of any one of the bills currently making their way through Congress, there will be more AWS spectrum and other spectrum that will be configured as FDD below 2.5 GHz.

Clearwire’s desire to move to LTE is the right decision, now it is a matter of timing. Can it find the funds needed to make the transition, how will it go about building out and converting its WiMAX customers to LTE, and how long will that take? Because it has so much spectrum available it could build out the TD-LTE network on an unused portion of its spectrum leaving the WiMAX network in place and then transition its WiMAX customers over to LTE without having to do it all at once. It will be tough, but if Clearwire is really in the wireless business for the long haul I think it will soon find a way to move forward.

Meanwhile Sprint’s transition to LTE (the FDD version) on its 1900-MHz spectrum will give it quicker access to many different types of LTE-capable devices, and since it is deploying a new set of cell site radios that can operate on multiple bands and using multiple technologies in the same box, its move to LTE should not be as painful. The fact that Sprint also now has an iPhone will help keep its customers around while it makes the transition but it needs additional funding in order to make this move. After it announced its move to LTE, Wall Street punished Sprint, which I don’t understand. It appears as though The Street is still in the “what have you done for me today” mode rather than looking at the long-term results of what Sprint is planning to implement

Even though I have for years written that I believe both Sprint and Clearwire had made the wrong technology decision, I understand that back then the decision was based mainly on Intel’s push in the WiMAX space and its willingness to invest big bucks in order to make WiMAX a success. Since then Intel has decided that WiMAX did not meet the promises made by the engineers and has gone quiet and stopped funding any WiMAX activity. I don’t have a clue what Intel spent over the years trying to push WiMAX around the world but it was heavily invested in the technology, at least for a while.

With both Sprint and Clearwire making the move to LTE it is clear that WiMAX will end up being relegated to point-to-point backhaul types of communications and that LTE will emerge as the prime 4G network of choice within the United States as it is in most of the other parts of the world. In all fairness to Clearwire and Sprint, it should be noted that when they were ready to roll out their broadband network LTE was not yet ready for prime time. The only other option would have been to deploy 3G technology and they would still be faced with having to migrate their networks to LTE regardless.

AT&T is joining in the LTE fun and has, at present, five markets up and running: Atlanta, Dallas, Houston, San Antonio, and Chicago. Since LTE data speeds are dependent on the amount of spectrum in which is it deployed, customers will notice really fast data rates in all of the cities except Chicago. In the other four cities AT&T is using 20 MHz of spectrum (10X10 MHz) but in Chicago it has only turned LTE on in a 10-MHz (5X5 MHz) portion of the spectrum. Still early reports are that data speeds, at least in the first four cities, are very good, but it is important to realize that the network is very lightly loaded at this point in time.

AT&T has an aggressive build schedule planned for 2012, and since it already has HSPA and HSPA+ deployed in most markets, the fallback from LTE to HSPA+ in particular will not be as severe in loss of data speeds as moving from LTE to CDMA EV-DO Rev A on the Verizon network. However, Verizon had a huge head start in build-out and is way ahead of AT&T when it comes to percentage of the population covered and I can attest to the fact that its LTE network is robust in many areas of the country today. I have yet to travel to a city in the United States where I have not had Verizon LTE coverage in my hotel room.

Speaking of Verizon, it has even brought LTE to Santa Barbara. Sitting at my desk and running my normal speed test software on my desktop system using Cox Cable, my incoming data rate is between 19 and 22 Mbps and my uplink data speed is between 3.7 and 3.9 Mbps (this is in the middle of the day when the heavy cable system users are in school or at work). I am very close to one of Verizon’s LTE sites (I can see it up the hill from me out my office window) and on my laptop I am getting downlink speeds of between 19.7 and 22.5 Mbps and upload speeds of 14.8 to 17.7 Mbps. Understand that this is a fixed location using an LG USB modem and I have virtual line-of-site to the cell site. However, I have no way of knowing how much data traffic is on this particular cell sector.

AT&T has not, to my knowledge, published any speed numbers it expects customers to experience on its network. Verizon has been consistent in its numbers telling customers that they can expect between 5 and 12 Mbps down and 2 to 5 Mbps up to the cell site. I have found that these numbers are good and reliable numbers as I have traveled the United States. I will be running tests on the AT&T network as well. It is interesting to me that there are already those who tout that AT&T’s LTE is faster than Verizon’s LTE. In reality, both are using 20 MHz of spectrum (except for AT&T’s Chicago deployment), and the same release of LTE software so they should both be offering about the same speeds.

However, within LTE there are many different parameters that can be set and that will affect performance. Some of these include Quality of Service (QoS), priority access, and being able to set up the number of data speeds in both the outbound and inbound directions. LTE supports three outbound data speeds (fastest, fast, and slower) depending on your proximity to the center of a cell sector, and two inbound data speeds. It is possible to adjust any and all of these parameters to increase or decrease data speeds. It is also difficult when conducting these types of tests to know how many other users are sharing the cell sector. If you have no one else within your cell sector you achieve the full total speed and capacity that can be delivered within that cell sector. However, if there are a number of other users within the same cell sector you will be sharing the total sector capacity (and therefore speed) with the other users. If three or four users are streaming videos down to their devices the results you will see during testing will be slower than if you are the only customer in a cell sector. Thus unless the network operator is working with you on the tests, you have no way of knowing how much other traffic is on that cell sector and could be affecting your test results.

T-Mobile is rolling out HSPA+ and has declared HSPA+ as 4G (as has AT&T), and in some cases the data speeds are pretty close to what you can get on LTE within the same amount of spectrum. T-Mobile only has spectrum at 1900 MHz and in the AWS-1 band so it is much more limited in being able to roll out LTE than the other network operators. I don’t expect to see any move toward LTE from T-Mobile in the foreseeable future. If the AT&T/T-Mobile merger does happen, then the issue is moot. However, if the merger does not happen, T-Mobile will have to step up big time at the next spectrum auctions in order to obtain more spectrum if it wants to add LTE to its existing 2G, 3G, and 4G offerings.

LightSquared

I saved this subject for last because there is a lot going on. For those who don’t know, LightSquared has asked the FCC to permit it to build an LTE network on its satellite spectrum, which is adjacent to the GPS band. There is a lot of concern about the potential for interference to GPS systems if LightSquared is permitted to build out its network. After tests proved there would be widespread interference, LightSquared offered to build its network using 10 MHz of spectrum that is further removed from the GPS frequencies BUT there are still a lot of unanswered questions about whether even this network will interfere with some if not all GPS units in the field.

LightSquared has pulled out all the stops to have this waiver approved. It is accusing the GPS industry of building shoddy products, it has said that if it does not win approval it will sue the FCC, and now it is saying that it has contracted with a company to build filters to be used to retrofit GPS receivers and it is willing to pay to have the filters installed on Federal Government GPS units, BUT it expects the GPS community and users to foot the bill (estimated to be about $400 million) to install the filters on other GPS devices that are effected.

Meanwhile, LightSquared continues to sign up wholesale customers, it continues its push based on how much its network will assist with new jobs and more money being spent in this down economy, and the fact that it will be offering competition to the existing 4G network operators. It appears to be hoping that its jobs, money, and competition strategy will overcome the objections of those who have the technical knowledge to understand the potential interference issues. The FCC recently ordered more tests between the proposed LightSquared system and the GPS community. From my perspective, I think that if the FCC folds and this system is authorized we will have a big problem on our hands.

LightSquared did not pay for this spectrum and when it bought the company that owned it, the FCC rules indicated that terrestrial stations could be used for fill-in of the satellite signals only and NOT for a nationwide terrestrial system.

It appears to me as though the decision to let LightSquared move forward with a terrestrial LTE system was based on politics and NOT on factual engineering data. In light of the current findings and until the new tests are run, even after they are run, it would be foolish to permit LightSquared to proceed with this system. If it wants to build a network for LTE for wholesale purposes there is spectrum available from Clearwire and there will soon be more spectrum coming up for auction that it could bid on that would work for its intended purpose.

Lots going on in the world of 4G, lots of devices coming soon, and it appears as though the world is waiting not so patiently for the iPhone 5 that is rumored to be LTE-capable and which is also rumored to be announced sometime after the first of the year. Until then there are many LTE choices today and there will be more over the coming months. LTE is solid, it works, it provides the best data speeds yet, and speeds will be getting better in future releases of the LTE specification. However, it is also prudent to understand that there will be times when a customer will fall back to 3G data levels so using applications and services that depend upon LTE to function properly does not necessarily make sense.

Andrew M. Seybold