Alereon: No Strings Attached

By Wes Simonds

November 30, 2004

The ultrawideband developer dreams of a world sans cables as it develops wireless connections with speeds rivaling FireWire.

Alereon CEO Eric Broockman wants to make the cable obsolete.

"Alereon's vision is to replace the complex tangle of wiring used to interconnect today's consumer electronics, computer peripheral, and mobile devices, with high-bandwidth, low-power, low-cost wireless links," Broockman recently said.

Like other personal area network (PAN) solution providers, Alereon is focusing aggressively on the perceived business opportunity in creating wireless links within a network node—involving, for instance, a tower CPU and its external monitor, mouse, keyboard and printer—in parallel to wireless links between nodes in a wireless LAN (WLAN).

This may sound familiar as the central business pitch behind Bluetooth, the earlier PAN standard now widely considered a non-starter. However, the underlying technology involved in this second wave of PAN solutions differs considerably from Bluetooth's. Alereon is using an ultrawideband technique, originally developed by the military and subsequently determined to have considerable private-sector appeal by engineers like those at Alereon.

Ultrawideband signals, unlike those of either Bluetooth or conventional WLAN technology, operate within a very wide range of spectrum—typically tens of thousands of megahertz wide, in fact. Their propagation is much poorer, limited typically to some thirty feet, but their performance is far greater, at least in initial prototypes.

In mid-November 2004, Alereon announced that it had achieved "the world's first over-the-air demonstration in compliance with the Multi-band OFDM Alliance (MBOA) specification... [displaying] speeds of 480Mbps and 320Mbps in their system lab this week."

That's literally hundreds of times faster than Bluetooth's maximum throughput (1-2 Mbits/sec), dozens of times faster than typical WLAN solutions based on 802.11b (11 Mbits/sec) and in fact competitive with USB 2.0 and the original release version of IEEE 1394, aka FireWire, the data storage standard commonly used for external hard drives and other high-throughput peripherals.

That's remarkable performance, and one might imagine it lends itself well to applications that go well beyond the PAN. Indeed, as Broockman suggests, UWB is seen as a candidate for handling any form of high-performance data throughput currently handled by cables, whether associated with the high tech market or not. Examples would include broadcasting DVD content from a laptop to a plasma television set or between camcorders/digital cameras and plasma televisions. Such is the throughput implied that one could conceivably upload dozens of MP3s from a portable player to a living room console in seconds simply by walking within range.

What's more, the interference issues commonly associated with conventional broadband wireless solutions, including Bluetooth, purportedly vanish with UWB. Because of its extraordinarily low power requirements (a tiny fraction of those required for a cellphone) and short range, UWB solutions are perceived likely to have no significant interference impact on existing wireless standards. For these reasons, the Federal Communications Commission (FCC) ruled in February 2002 that UWB devices could operate from 3.1GHz to 10.6GHz without requiring a license.

Such technology is bound to have an eager market. Unfortunately, Alereon's process in bringing the technology to the public has been stalled over squabbles about the definition of the nascent UWB standard 802.15.3a. In a story familiar to anyone who's followed the general evolution of broadband wireless technologies for the last four years, there's been an IEEE collision between competing factions. Intel leads one group, known as the Multi-Band OFDM Alliance (MBOA); Freescale (formerly known as Motorola in this context) leads the UWB Forum.

One might imagine from the ongoing delays for standard ratification that the struggles have been evenly balanced. Broockman disagrees.

"Most industry insiders know that outside of the IEEE process the battle has been over for some time," he said. "An impressive array of companies including Intel, Microsoft, Nokia, HP, Sony, Sharp, Samsung, Toshiba and LG are members of the MBOA. In fact, nine of the top 10 semiconductor companies are MBOA members (#10, TSMC, is uncommitted) and 100 percent of the leading consumer electronics companies back the MBOA proposal. Additionally, both the Wireless USB Promoters Group and WiMedia have endorsed the MBOA approach."

What does the post-ratification future hold for this technology? Some see UWB technologies as ultimately colliding with WLAN technologies in a market confrontation as UWB scales up in performance and, presumably, extends its range.

Broockman doesn't. He feels the PAN and WLAN markets will instead continue to develop in parallel.

"WiFi is wireless Ethernet," he said, "while UWB will enable wireless USB 2.0 and 1394. These are complimentary technologies for good reasons and will remain so in their wireless incarnations."

Nor does he feel the future WLAN standard 802.11n, with projected performance in the 200 Mbits/sec. range, can threaten UWB technologies in the PAN market.

"802.11n holds the promise of extending the ubiquitous IP network to yet another set of applications, but 802.11n will face difficult challenges - including security concerns, interference with other wireless protocols in the same spectrum and bandwidth concerns - especially for mobile applications. UWB radios are over ten times more power efficient per megabyte of data transferred than 802.11n as well as being much faster. This has nothing to do with Moore's law. It is simply physics. This is a very compelling advantage. "

In other words, we'll have different standards in place simultaneously, undeterred by interference issues and solving different problems with different approaches. That's arguably a utopian perspective, but given the relevant technological issues, it may be a realistic one as well.



Comment and Contribute
(Maximum characters: 1200). You have
characters left.