Almost every day I see news of further developments in advancing the speed and performance of home networks. Every IC, box, and software vendor who has any sort of ambition or investment in this market touts their approach as being the best and most likely to succeed. Who can blame them? This market is big and will be getting bigger as the home becomes more of a data and multimedia center with many access points.
The three contenders are wireless (based on 802.11, most likely); wired (using in-wall coax or CAT5 cabling, generally); and the existing AC power line. But while there is a standard for the latter and available products, but frankly, I don't see it gaining more than a fairly small portion of the market.
Why not? I'm not going to give you a feature-by-feature comparative table. Instead, I'll give you my "high-altitude" overview reasoning. Yes, the power line is everywhere in the home (that's its main virtue) but getting any sort of decent, consistent low-latency performance for near-real-time streaming, or even low-enough BER so packets don't have to resent and resent, is very difficult. It's a lousy electrical environment, period. It takes a lot of relatively expensive and power-hungry circuitry to achieve data rates that are even an order-of-magnitude lower than what a much less costly wireless network can provide.
Further, wireless-system performance is not standing still, and is increasing at a fast pace, yet with only a modest increase in system cost, complexity, and power. Some of this is because the wireless vendors are taking advantage of high-volume, low-cost RF technologies and components developed for handsets and base stations. Technically, AC line-based systems are having a very, very tough time keeping up with the speed advances of wireless and the demands of the home market. It takes a lot of signal processing, filtering, and protection circuitry to make that AC line system work even reasonably well under "typical" conditions, and then someone next door plugs in a noisy, brushed-motor electric drill, which feeds hash back onto the line, and performance really droops.
As for wired systems using cables in the wall, if the cabling can be put in place during construction, there is no doubt that it yields a fairly cost-effective and reliable way to set up a high-performance network. In fact, it can be used as a house-wide backbone supplemented by wireless distribution within each room, for a system which has the virtues of high speed, low error, and no tethers. But in general, it's a cost-effective solution only where the cabling does not have to be retrofitted.
So I am betting on wireless systems as the overall winner, despite the fact that some locales, such as those with thicker walls or RF anomalies, may not work well with wireless. In second place, I'll put in-wall cabled systems. Lagging far behind will be AC-line based systems, despite their legitimate claim of ease of physical installation.
There some personal irony here. Way ,way back in the day when I was in school (we're talking ancient history here), one of my fellow students had a modest research grant from now long-gone minicomputer giant Digital Equipment Company (DEC) to develop a system that would push data at 1200 baud over the power line. He worked hard, but ultimately was not very successful. It took so much processing and filtering and other tricks to make that system work, it simply wasn't worth it.
Sure, the requisite electronics at each end are now far more capable and sophisticated, so 1200 baud would be relatively easy. But that's the problem: 1200 baud is not an acceptable goal these days. The market goal posts keep moving, and they are always several very big steps beyond what AC line-based systems can effectively provide, in my view. I don't think that is going to change.♦