Over the years, we've written many articles about what we term "whole home networking": networks capable of carrying high-quality digital video, audio, telephony and data throughout the home. We have long believed that all existing networking technologies are "under the bar" -- they lack the QoS mechanisms required to support time-critical applications like video and voice, and lack the application-level data rate for high-definition television.
Now, many new technologies are coming to market nearly simultaneously. Some are wireless, while others operate over existing wiring (electrical, telephone or coaxial cables) in the home. These technologies all seem to be "over the bar": all have QoS, and all claim to have high enough data rates to support HDTV.
We think we're at the "start of the end game" for home networking. Far too many technologies are competing for the networked home. By the end of 2007, it's likely to be clear which will win the mass market -- possibly a single technology, more likely a combination of several technologies.
Video Applications and Requirements
Until now, most home networking applications have been data-oriented -- most often, sharing a single internet connection among several PCs in the home. New video applications promise to change the game and add new requirements.
Many emerging consumer applications require networking technologies capable of moving video around the home. These include:
These emerging consumer video applications establish key requirements for networking technologies:
Many emerging technologies are being positioned as candidates to satisfy the requirements for whole home networking. Some are already on the market in low volumes; most others will reach the market within the next year. Each was designed to be carried over a particular medium -- existing wiring (powerline, coaxial cable, phone line, and the air (wireless). Each has advantages and disadvantages.
Phoneline technologies are designed to operate over the existing telephone wiring in the home. Early phoneline technologies--HomePNA 1.1--were adopted by some telephone companies to carry data services between the home gateways and home PCs. The most recent technology--HomePNA 3.0--is targeted to video networking, and several products have come to market designed for this application.
HomePNA 3.0 is designed to operate over wiring that also supports traditional analog telephony and DSL. Telephone companies like HomePNA since they control the telephone wires in the home.
Home phone wiring is often old, and sometimes poorly installed. North American homes typically have phone outlets in many rooms, but rarely located close to TV sets. Outside North America, homes typically have fewer telephone outlets.
Cost reduction is uncertain. Although HomePNA 3.0 is promoted as a standard, it was not developed through an open standards process but rather through an industry consortium. Only one semiconductor company appears committed to the technology.
Powerline networking technologies are designed to operate over the existing electrical wiring in the home. We found an earlier powerline technology--HomePlug 1.0--to be quite satisfactory and we use it in our home audio system (see "For More Information" below for links to our earlier articles).
Several powerline technologies are competing for future applications, most notably DS2 and HomePlug AV. Products based on the latest DS2 chips have been adopted by telephone companies for networking between IPTV gateways and set-top boxes. Products based on HomePlug AV should be on the market later in 2006.
Since electrical outlets are ubiquitous in homes, powerline has long been viewed as the ideal "no new wiring" technology for tethered devices -- build the chips into the devices, plug them in, and they're networked.
But the electrical environment in the home is very noisy, and differs from home to home. The path to low cost is not clear. None of the powerline technologies has been developed through an open standards process, although multiple vendors are participating in the HomePlug industry alliance and DS2 is working through ETSI and the IEEE.
In the past, powerline technologies have not been marketed effectively. No major semiconductor companies have produced powerline chips. Intel's recent enthusiasm for powerline networking and its assumption of a leadership position in the HomePlug Alliance is encouraging.
Coaxial cable typically runs between the TV sets in the house, installed by the cable or satellite provider. Several companies have developed and released semiconductor technologies for high-speed networking over coaxial cable, including Entropic Technologies and Coaxsys. Several telephone companies have announced deployments of coax networking to distribute IPTV between the home gateway and IPTV set top boxes.
Since coax is designed to carry multi-channel analog television, it can carry high bandwidth and is comparatively "clean" -- if television looks good, high-speed networking will probably work fine.
Since coax networking is a nearly-ideal solution for video service providers, many companies are competing to provide the technology. In addition to the companies whose technologies were designed for coax, several technologies designed for other media are being adapted for coax; these include DS2 (originally for powerline) and HPNA 3.0 (originally for phoneline and now being repositioned for a mixture of coax and phoneline). We have been told that at CES one flavor of UWB (provided by Pulse~Link) was demonstrated operating over coaxial cable at 320 Mbps.
Cost reduction is questionable. Most coax technologies have been developed by individual companies. While some are being promoted through industry consortia, there are no open standards.
Coax is a good medium for networking between TVs. Since coax rarely appears in other home locations, it is not as useful for other applications.
Over the Air (Wireless)
The next generation of "Wi-Fi"-- 802.11n -- provides both QoS and the data rates required for video; the addition of MIMO suggests that it may also provide the necessary range. After some delays last year, 802.11n appears to be back on track (see our article on CES elsewhere in this issue). "Draft n" products are likely to appear on the market later in 2006, with fully certified products by mid-2007.
802.11n is very promising. Our tests of early MIMO-based products suggest that both the rate and range may well work for video. But the use of wireless over long distances for trouble-free video networking is unproven until real products are tested in real homes.
UWB offers another solution for short-range wireless video. The Wimedia flavor of UWB will appear on the market later this year, and some vendors are talking about using it to connect set top boxes to portable TVs.
The End Game
The simultaneous introduction of all these new technologies is almost certain to engender a great deal of confusion among consumers and the lay press. But we believe that the confusion will die down as the end game becomes clear.
In the late 19th century there were several competing forms of electricity (AC and DC), each with its advocates -- AC won and DC faded out for home use. Cylinders and disks were both used as recording media competing vigorously for the consumer's ear. Everybody remembers the VHS/Betamax war for VCRs during the 1980s.
We don't think the home networking market can sustain so many different approaches, and we believe that video service providers and equipment manufacturers will decide which technologies to back and which to let fade away.
There appear to be several broad scenarios: one is "wireless-only" and several include a wired "backbone".
Wi-Fi has been the big winner so far in "no new wires" home networking, and its leaders expect the new 802.11n "flavor" to play the leading role in the end game. They believe that all forms of wired technologies will play only a small specialized role; 802.11n wireless will dominate and will be built into most home devices, whether tethered or portable.
Powerline advocates see a very different endgame. They believe powerline networking will provide the "home backbone". Powerline chips will be built into all "tethered" devices, which will be networked automatically and effortlessly as soon as they are plugged into an electrical outlet. In their favored scenario, wireless will be used only for specialized portable applications, such as notebook PCs, portable Wi-Fi phones, and portable TVs.
Coax advocates see a different "backbone" scenario. Since they concede that coax is not as ubiquitous as electrical wiring, they see the need for wireless as well. Nevertheless, they see coax networking used as the underlying backbone for video distribution throughout the home--perhaps the most sensitive application--with wireless used to provide connections from coax outlets near TVs to other locations and devices in the home.
There are other possible scenarios as well. UWB enthusiasts think UWB will be used for high-speed communications between high-definition set top boxes and TVs, augmenting or replacing the current analog and digital (DVI or HDMI) cables. HPNA 3.0 enthusiasts think it will be used to extend the backbone network using both phonelines and powerlines.
Until now, consumers have made the key decisions on home networking. Video service providers may play a major role in deciding how the end game plays out. Telephone companies need an easy and reliable way to transport IPTV streams from DSL or fiber gateways to set top boxes. All video service providers are looking at multi-room PVRs, and need a way to get from the central PVR to the set top boxes and/or TV sets. If these companies decide to provide whole home networking as part of their service offerings, their evaluations will determine which technologies to incorporate in their gateways and set top boxes.
By the end of this year, we should know how closely each of the new technologies comes to meeting the requirements for whole home networking. By the end of 2007, the playing out of the end game should start coming into focus.
For More Information
We have written extensively about all of these technologies. For more information,