HomePNA comes screaming back

The Home Phoneline Networking Alliance has had a tough few years. First lauded for its ability to produce a solid specification in an orderly and timely fashion, HPNA's Version 1.0 products flopped in the retail market. Then came Version 2.0 (noticeably late), which increased speeds from 1.6M bit/sec to 10M bit/sec. But it was too little too late. By then, 802.11b wireless was all the rage, and 802.11a (rated 54M bit/sec) and HomePlug products (rated 14M bit/sec) were on the way. Why struggle to configure a home network around a few poorly placed phone jacks when you can use power outlets and wireless to work anywhere? Never mind that HomePNA technology has always been cheap and reliable. When home network technologies first began vying for market share three years ago, the ability to replace 10M bit/sec wired Ethernet was all that mattered. That's partly why 802.11b trounced HomeRF so soundly.

Today, the big news is that HomePNA may actually win the performance race after all. Version 3.0 of the home phone line specification, announced earlier this month, offers 128M bit/sec speeds. Using extensions vendors can build into their products, speeds can reach 240M bit/sec.

Version 3.0 is the result of a technology developed jointly by Broadcom and CopperGate Communications, which had each been vying to win the spec. Word is, the alliance reached a stalemate on which technology to choose. CopperGate was strongest on the media access control, and Broadcom was strongest on the physical layer. In the end, the two companies teamed and submitted a joint proposal.

"The combination of these best-of-breed technologies creates a real powerhouse," says Dave Thomasson, HomePNA spokesperson and CopperGate vice president of marketing. "With these speeds, you can go into an environment that doesn't support HPNA well and still get good throughput. With QoS, you'll have discrete channels for each data stream. When two packets collide, they'll each randomly generate a number, and the higher one will go first. With voice, audio and video, QoS is critical. You can't interrupt the data stream and have jitter, or visual or audible problems."

The spec will be finalized in November, followed by field trials in 500 homes. If all goes well, the first bridges and adapters should be available in summer 2003. You can also expect to see HomePNA 2.0 and 3.0 more frequently built into residential gateways (namely, 2Wire's Home Portal) offered by broadband service providers, either as a HomePNA or mix of wireless and HomePNA home network package.

Thomasson sees home networks evolving into a hybrid of technologies, with wireless delivering the signal to the final device, but carried through the home using a HomePNA backbone. The HomePlug folks are vying for the same space, but the technology is relatively immature and much slower, assuming HPNA 3.0 isn't delayed. Still, Thomasson says, "It's such a huge market, and we've seen such a tiny piece of it so far. When the consciousness of the average consumer is raised and people want this technology en masse, there will be solutions for everybody.

"It's very exciting to see this stuff come together after all this time," he adds. "Everything takes longer than you think it will, and always costs more."

RELATED LINKS