BBH Report Home Page
November 16, 2003 Provided by System Dynamics Inc.

What's Next for Wireless Networking

802.11g--the newest version of Wi-Fi--is a fully-approved standard and lots of products are now on the market. It provides much higher speed than the earlier 802.11b, but it's certainly not the ultimate wireless networking product. It isn't the solution for "whole home networking" -- carrying high-quality digital video and audio and telephony as well as data -- since it lacks QoS and is substantially degraded by walls and floors. And it doesn't work fast enough for high-definition TV, which is gaining increasing market share.

A recent announcement of a deal between Appairent and Pulse~LINK, two companies working on next-generation wireless technologies, got us interested in whether these would any better at carrying HDTV throughout a home.

Several wireless technologies are designed to handle HDTV; some will be on the market soon, while other are several years away. Here's a summary of three technologies we're following:

  • 802.11n - Confusingly, 802.11n is the follow-on standard to 802.11g which followed 802.11b. 802.11n is aiming at 100 Mbps throughput -- a four-time (or more) improvement on 802.11g, and enough to carry multiple simultaneous channels of HDTV. The standards group has just been formed, and its charter anticipates completing the standard in mid-2006. We hope 802.11n will do better than 802.11g in penetrating the walls and floors of homes, but suspect it won't since this isn't stated as a requirement in the evaluation process for potential technologies. The 802.11 group has always seemed ambivalent as to whether its standards are for networking in offices or in homes; most of its technology evaluations have focused on office environments.
  • 802.15.3 - This is a recently-approved IEEE and ANSI standard for high-speed "personal" area networking (PAN). While the 802.11 group has focused on "local" area networks (LAN) with a range of 100 meters, 802.15 deals with personal networks with a range of "at least" 10 meters. Designed to operate in the same 2.4 GHz band as 802.11g, 802.15.3 promises considerably higher throughput and includes QoS. Some advocates claim it will do better carrying HDTV through walls and floors.
  • 802.15.3a - This group is working to develop a follow-on standard to 802.15.3, using the same MAC layer with built-in QoS, but with a different PHY layer based on ultra wideband (UWB). It aims at 100 Mbps throughput--the same as 802.11n. While positioned as a PAN, advocates say it will be competitive with 802.11n in the home environment and will be completed much earlier.

Appairent -- Why the MAC is important

We talked with Tom McGovern, the CEO of Appairent, a spin-out from Kodak. He explained that 802.15 provides a different MAC and PHY layer for each "sub-dot" (.1, .2 etc). 802.11 has always used the same MAC, and that has gotten 802.11 into trouble: a MAC designed for 2 Mbps doesn't work very efficiently at 54 Mbps (802.11a and 802.11g) and won't work at all for 100 Mbps throughout.

Appairent is very active in 802.15, and its CTO, Dr. Robert F. Heile, is chairman of IEEE 802.15.3a. While most companies involved with wireless networking technologies have concentrated on the physical or PHY layer, Appairent chose to focus its energies on developing a highly-efficient MAC and upper layer stack for 802.15.3 and .3a. This is fully operational and is available today as part of a complete 802.15.3 evaluation platform.

Tom told us that 802.15.3 products "will cover the average household with 42-45 Mbps". He said the 802.15.3 MAC is far more efficient than 802.11; while the 802.11 MAC is contention-based, 802.15.3 is based on TDMA. Although 802.11g and 802.15.3 run at almost the identical maximum physical speed (54 and 55 Mbps, respectively), the efficient MAC layer in 802.15.3 provides twice the throughput of 802.11g.

Work is well under way to create the ASIC implementation, and Appairent and its customers plan to bring products to market during 2004. We're looking forward to testing one of these to see how it does in our house.

Pulse~LINK -- The Benefits of Ultra Wideband

Pulse~LINK is a leader in developing ultra wideband (UWB) communications technologies; we reported previously on their work on UWB over coax ( Following Pulse~LINK's announcement of a deal to license Appairent's MAC to create an ultra wideband (UWB) wireless LAN system, we interviewed John Santhoff (founder and CTO) and Bruce Watkins (President and COO) to get an update on the role of UWB in wireless networking.

John said he had attended many 802.11 and 802.15 meetings -- these are usually held jointly and members have reciprocal voting rights. At a recent meeting, the consumer electronics representatives said they wanted to cover a full home; a range of 60 to 80 feet, going through walls and floors, should be a requirement.

John said Pulse~LINK's technology was capable of going 100 meters and they will couple it with Appairent's MAC and higher layers to create a LAN solution now.

They think UWB will prove "much more robust" for wireless networking than 802.15.3 or 802.11, which operate over comparatively narrow 20 MHz bands. Since UWB uses a very narrow pulse that goes across 1 GHz or more, it is much more resistant to the "frequency-selective fading" that afflicts the narrow-band technologies.

They think UWB will be used two ways in the home. The first is to interconnect nearby devices (PC equipment or consumer electronics) by communicating over the power line -- just plug the devices into the same power strip and they're talking with each other. The second is using UWB over wireless to reach mobile devices.

Consumer products based on Pulse~LINK's UWB technologies should reach the market in 2005.

The challenge for all 802.15 technologies will be the 802.11/Wi-Fi bandwagon. With the enormous marketing power behind Wi-Fi and its rapidly-growing home penetration, it isn't clear than any new wireless technology--even if better--can succeed in the home networking market. The best opportunity comes from the likelihood that 802.11n will not come to market for three years or more, and will include only limited--if any--back-compatibility with 802.11g.

We will continue to watch these emerging wireless networking technologies, and other "no new wires" technologies based on existing electrical, telephone and coaxial cable wiring.

( ) ( ) ( ) ( ) ( )