Broadcom: We're not confident about WiMAX in the U.S. any more

It would appear that U.S. companies are ready to abandon WiMAX.

At the end of July, Sprint made the bombshell announcement that it had entered a 15-year partnership with LightSquared to share its network infrastructure with the new network, and that it would essentially be receiving a free LTE network if it agreed to build it on top of its existent towers. Sprint's WiMAX partner Clearwire, has laid out no further plans for its WiMAX expansion, and also has committed to incorporating LTE into its network, and eventually upgrading to LTE-Advanced several years down the road.

This week, we asked wireless chip manufacturer Broadcom about the situation, and received the answer that you could have guessed. Yes, WiMAX has been abandoned.

Last year, Broadcom acquired a company called Beceem that was a leader in chipsets for 4G WiMAX and LTE. Michael Hurlston, General Manager and Senior Vice President of the Wireless LAN Business Unit at Broadcom told us that the Beceem team is more or less done with WiMAX.

"The WiMAX business is getting isolated to certain geographies," said Hurlston. "It's doing well in, for example, India, and doing relatively well in Japan, and somewhat in the Eastern Bloc, but in the U.S., I would say that it's not doing particularly well. The reason for that is that LTE, the new cellular standard, is coming in and offering what WiMAX purports to offer, which is very wide area coverage at high speeds."

"Sprint was a big proponent of WiMAX when they got involved with Clearwire, but that effort has largely slowed," Hurlston continued. "So as it pertains to the United States, at least from our perspective, WiMAX is eddying, it doesn't seem to be making a lot of progress and we're not confident that ultimately it will be something that's going to be very interesting, despite us buying a company that was focused on WiMAX. I think that we're re-tasking that particular chip team to focus on LTE."

This is a far cry from the climate two years ago, when WiMAX advocates touted the mobile broadband technology as the answer to rural America's broadband needs.

For this type of connectivity, Broadcom has put its support behind the IEEE standard 802.11af, which is essentially wide area Wi-Fi that exists in the "white spaces" of the spectrum, or the areas where radio communication is allocated to TV broadcasters, but which are vacant. The available spaces vary from region to region, so the standard includes "cognitive radio" functionality, where the end user's device will intelligently determine where the free spaces are, and use them.

"The TV white space spectrum has opened up in the 800-900MHz range, and the 802.11 task group is looking to take advantage of that spectrum and make use of a much longer range wireless LAN solution, typically to serve metropolitan wi-fi, but also to serve rural areas, which the government mandated," Hurlston said. "When you go down in radio frequency, the signal's range increases, and right now 802.11 works in the 2.4GHz or 5GHz range, and that's part of its coverage limitation. When we start talking about 800MHz, what you get into is a much larger coverage range…with 802.11af, we're talking about a maximum range of about five miles from tower to access point."

WiMAX operates in the microwave frequencies, meaning it's got a super high frequency, but a small range and susceptibility to line-of-sight interference issues. Clearwire deployed its cell sites at intervals of about 1.5 miles.

802.11af has got a long coverage range, so the base station infrastructure for a network using the standard would be considerably cheaper than WiMAX. The tradeoff, however, is that the bandwidth is relatively low.

"In Wireless LAN we talk about bandwidths of 50 Mbps," said Hurlston. "What 802.11af can deliver is between 5-10 Mbps, and the further out from the base station you go, it decreases. At maximum distance, you might be looking at about 1 Mbps. So there's definitely a range vs. throughput tradeoff that happens. To a certain extent, there is channel bonding, but the reality of it is that you're dealing with much narrower, 5MHz channels, and that affects throughput."

So for purposes of bringing connectivity to the underserved areas of the United States, a network that can reach more users while costing less money is preferable to high bandwidth at a high cost.