LTE roaming likely a long way off
Several wireless carriers made announcements about their LTE plans over the past few weeks. We saw updates from Verizon Wireless (NYSE:VZ), AT&T Mobility (NYSE:T) and, of course, MetroPCS (NASDAQ:PCS), which launched its LTE service in Las Vegas, giving the flat-rate, prepaid carrier claim on the first commercial LTE launch in the country.
All of this is great news for almost everybody--carriers, vendors and consumers--that want to take advantage of faster speeds and new applications. However, there remain a number of knotty issues--the thorniest of which is LTE roaming.
Consider this: For an LTE handset from AT&T to roam among the viable LTE networks in the United States over the next few years (assuming Clearwire (NASDAQ:CLWR) switches to LTE), it likely will have to support--at the least--all of these radio frequency bands:
- 700 MHz
- 850 MHz
- 1700 MHz
- 1900 MHz
- 2100 MHz
- 2.5 GHz
Admittedly, this is a much more complicated issue in the United States, where LTE spectrum allocations are more fragmented, than it is in Europe (many European countries are coalescing around 2.6 GHz as a primary band for LTE). But it is still a worldwide issue that deserves a great deal of attention. Unfortunately, for a variety of reasons, I think it likely will be years before most carriers strike LTE roaming deals. Until that happens, customers will not be able to fully take advantage of what is emerging as a truly global mobile standard.
So what are the key barriers to LTE roaming? Surprisingly, the barriers are less about technical issues than commercial ones. Dan Warren, director of technology at the GSM Association, told me that in late 2008, the group started the Next Generation Roaming and Interoperability initiative to devise technical standards for LTE roaming. The GSMA completed its work in March, Warren said. "We've put all of the tools on the table for people to pick them up and use them," he said.
Warren said he thinks carriers will start LTE roaming agreements by the end of 2011. A key determining factor, he said, will be how quickly device makers release devices with support for multiple LTE bands, as well as how quickly spectrum positions get harmonized. Still, Warren conceded, operators probably won't make roaming agreements a top priority.
"I think it will be natural for operators to want to establish their own network before they invite customers from other networks or roam onto them," he said, emphasizing that roaming should come sooner rather than later. "From a consumer point of view, that has to be fairly short term."
Others are less optimistic. Phil Marshall, an analyst at Tolaga Research, said he expects large-scale LTE roaming agreements by 2013 or 2014. It's a commercial issue, he said. Carriers--especially AT&T and Verizon--likely will build out parallel footprints before even thinking about roaming.
Roaming also will depend significantly on how many LTE subscribers there are. According to an August report from Juniper Research, only 1 in 20 subscribers worldwide will be using LTE by 2015. Although this could change, it fits with other forecasts from Maravedis and WiseHarbor Research. "You need to have an adequate market scale of LTE to justify that investment in multi-mode LTE capabilities," Marshall said. "Device makers want to know there's an adequate amount of volumes."
Steve Brown, the senior director of product management of RF products at chipmaker Qualcomm (NASDAQ:QCOM), said his company's latest chips have support for multiple LTE bands, but acknowledged the headache of engineering the new chips. Once European bands and those in Asia and other parts of the world are thrown into the mix, he added, it becomes "a pretty big matrix."
Both Marshall and Brown said that software-defined radios will help solve part of the problem. Still, RF components will have to be miniaturized on a massive scale for LTE roaming to take off, and that probably won't happen until there are large enough volumes.
I don't doubt that LTE users eventually will be able to roam. However, LTE roaming likely will be delayed by both carriers' desire to build out their own networks and device makers' reticence to engineer multi-mode LTE devices. --Phil