At the Mobile World Congress trade show in Barcelona today, the next-generation 4G wireless service finally got some respect, with AT&T saying it will likely deploy the Long Term Evolution (LTE) standard in 2011 rather than in 2012 and Verizon choosing vendors for its upcoming LTE rollout.
Verizon has chosen Ericsson and Alcatel-Lucent (s ALU) as the initial vendors for the LTE rollout and is reporting speeds of between 50 and 60 Mbps downlink in test markets. This leaves other vendors who participated in the trials — Nortel, Motorola and Nokia-Siemens — out in the cold so far. As far as speeds go, once the carrier puts its millions of subscribers on the network, those speeds will change, but it’s still going to be better than the current 3G when it comes to delivering data.
And data’s what Verizon’s LTE network will be for, according to Verizon CTO Dick Lynch. When I asked him about LTE handset deployments, he said there will be a few handsets and laptop cards with LTE about a year after the 2010 LTE network launch, but he stressed that voice is going to be carried over the 3G network well into the next decade — LTE will be for devices. In the future, consumers will have 8-10 LTE-enabled devices that “the consumer won’t even know has broadband,” he said.
Lynch talked about LTE on cameras to automatically upload pictures once a memory card is full, and LTE in cars to deliver information and entertainment. Unfortunately he didn’t have time to tell me how or who Verizon would charge for access to the LTE network in these scenarios. For the world to experience the joys of ubiquitous broadband on everyday consumer devices like cameras, personal navigation devices and eReaders, prices for LTE chips need to cost less than $10 or (better yet) $5.
Verizon also needs to have a business model that doesn’t require pricey network connection subscriptions for the end consumer. Is Nikon going to pony up a network fee just so I can upload pics from my camera? Somehow I doubt it, which means Verizon, Nikon or some independent vendor is going to have to come up with a way to recoup those costs. Perhaps a membership fee to a web site where the pics are stored?
Although the details are still up in the air, Lynch is promoting Verizon’s open network for LTE-enabled devices, and he predicts an exabyte of data traveling over Verizon’s LTE network in 2013. For more details, check out his slide presentation.
Hopefully, this will evolve in such a manner so that you have a single network subscription which can be accessed from multiple devices. However, something tells me that affordability is not their #1 priority. After all, their network architecture dictates diminishing QoS as the network grows which can only result in reduced revenue-per-account and higher OA&M. I doubt there’s much motivation to “give it away”.
There are definitely going to be field tests from LTE that are a nice proof of concept, just like the DoCoMo tests from a few years back where they hit over 40Gbps wirelessly. The problem is dispersion of sites and, more specifically, network congestion. The more people on the network the more people pulling from the same bonded T1’s in the bottom of the cell site. This translates into slower speed per user. THINK CABLE INTERNET, the more people on your node the slower the network is at the point of use.
For LTE to be anything more than an evolutionary increase in data transmission is to wish that carriers build network infrastructure consistent with the idea of a data revolution. LTE has the potential to revolutionize the way people access information and content by reducing the barriers between mobile workstations and stationary ones.
If LTE isn’t rolled out in a manner that allows enough cell sites so that every user can get a big enough piece of the pie, what we will have is an incremental increase over HSPA. A simple incremental increase would be remarkably narrow-minded and not in the best interests of the carriers or the consumers.
-Josh
if LTE has limited coverage, how can you use it in embedded devices?
Stacey, you quoted Verizon on the element that “voice is going to be carried over the 3G network well into the next decade – LTE will be for devices.”
Do you know if the same think is true for WiMax?
I call BS. The “network” includes subscriber devices. You can’t launch a network without subscriber devices. If there are no retail-available devices, you have not launched your network. The real national launch of LTE in the US will be 2011 or more likely 2012.
@Dee
Verizon is twisting the truth. Their entire network is based on the CDMA2000 standard. However that standard uses different radios to carry voice and data. The voice radios should actually be classified as 2G, while only the data radios can be classified as 3G. Since it is all “one network” they are claiming that the whole thing is 3G, but that isn’t really true. Sprint’s has a CDMA2000 network as well. The same thing goes for GSM/UMTS operators like AT&T and T-Mobile — circuit-switched radio for voice, packet-switched for data.
WiMAX only has one kind of radio and it is only for data. Any voice application on WiMAX will have to be VoIP.
Ok, this is the last time I am going to try and post this comment (it went into a vortex 3 times).
It is funny that all bloggers are only focusing on the LTE aspect of the deployment and talking about Ericsson and ALU. Remember LTE is only the packet-core and RAN part. Verizon plans to deploy IMS on top, and for that they have selected ALU as well as NSN (see his slides). The new apps will reside above the IMS layer, not the LTE later.
Arjun, sorry about the vortex, and thanks for pointing out the IMS layer wins.