On-Chip Challenges
Now that you get the gist of how beamforming works, you’re probably wondering why the technology never went mainstream. After all, in comparison, trying to optimize signal strength with several antenna jutting from a conventional 802.11n access point is a joke. These multiple antennas are, in a way, glorified rabbit ears. Even if you spend the time to fiddle with them and get what seems to be the best throughput in a given spot with the antennas set in just such a way, what happens if you have to move the access point or the end client? What if you add a second or third client? It’s chaos. The fact is that proper signal optimization with current-gen products is futile.
Why hasn’t intelligent beamforming, which has the ability to sense optimal phasing and orient beams for multiple clients, been widely adopted? It’s a mystery—probably another one of those “we as an industry are still in the process of discussing various blah blah blah” things.
A skeptic might suggest that on-chip beamforming hasn’t taken off because it’s sounds better on paper than it is in real life. We know that, in theory, beamforming should save power. You only need to boost the signal in a certain direction and drop power for any signals that don’t assist that beam. The problem here is that when you’re dealing with omnidirectional antennas, there’s only so much control you can have over your beams.
For an intriguing illustration, check out Falstad’s Antenna Applet and be sure to choose Broadside Array from the top pull-down menu. You can increase antenna counts, play with the distances between them, and modify signal strengths. As you’ll see, with two omnidirectional antennas you never get away from having a lot of beams, and therefore energy expended in unneeded directions (these unwanted beams are often called backlobes). Naturally, if you have beams going off in stray directions, these can cause co-channel interference and impede the signal you actually do want.
It seems likely that next-gen 802.11n will incorporate implicit and explicit beamforming at some point, as there are very few technical or cost barriers. However, which approach will vendors integrate? And we haven’t even scratched the surface on options. For example, there are three sub-types of explicit beamforming. So if there’s one culprit behind the lack of beamforming adoption, concerns over interoperability is probably to blame. For those of who find yourselves thinking, “Come on! I don’t care about 100% interoperability. I just want crazy good wireless performance in my space,” keep reading.
