« Home Ecophobia | Main | The Inappropriate Advert »

April 14, 2006

802.11n: What they won't tell you

Noseguy_1 While we've touched on this before, we wanted to dig a bit deeper. We continue to get bombarded with questions and snide remarks about 802.11n and how it effectively eliminates the Ruckus around multimedia over Wi-Fi.

It doesn't.

Most folks still believe that 802.11n will solve the problem(s) of sending video and real-time multimedia over Wi-Fi.

It won't.

But don't get us wrong....802.11n will be fantastic, and we'll be one of the first ones out there spewing 802.11n in all its glory. But caveat emptor.

802.11n could actually be more prone to interference and performance fluctuations than previous Wi-Fi technologies. Here's why (strap yourself in...this will get controversial, quickly).

802.11n splits a data frame into multiple pieces. It then transmits these pieces in parallel using multiple radios that are attached to multiple antennas (this is where MIMO - multiple in, multiple out - comes from). The receiving device also uses multiple antennas, radios and processing to recover the streams. This technique, known as spatial multiplexing, relies on different RF "propagation" paths.

Each path, though, has a slightly different delay between the time the signal is transmitted and received. If the paths between the sender and receiver are similar, then the transmitter reduces the bit rate by reducing the number of "spatial" streams. Basically 802.11n relies on this delay to work.

The benefit of doing all this is to increase performance and throughput. But it also increases the number of things that can go wrong with any of those streams. While three antennas give you a better chance to get a better RF signal, there's still interference with which to contend. And these (typically) omni antennas that are used just blast out signals from virtually the same vantage point - scattering the signals everywhere. Being able to control where the signals go and when would be ideal (we can do that with smart antennas).

Significantly more complex than the current 802.11 a/b/g Wi-Fi standard, 802.11n also introduces a new range of parameters that need to be controlled. These include things like:

  1. Channel bandwidth (20 MHz vs. 40 MHz)
  2. The number of spatial streams (1,2,3 or 4)
  3. Space-time block coding options (don't ask)
  4. Beamforming options
  5. Variable guard intervals (don't ask again)
  6. Frame aggregation
  7. Block acknowledgment(s)

All this points to the need for control algorithms that can intelligently and dynamically tune these parameters to ensure predictable performance.

In marketing spew: "an application-focused control framework is essential for dealing with difficult real-time optimization problems and will prove to be a vital technology for maximizing 802.11n performance" (yeah I don't know what that means either but I literally wrote it down when one of our founders said it to me).

What he was trying to say was that 802.11n needs stabilizing. Our stuff does exactly this and can be layered on top of new 802.11n chips to stabilize the link so performance can be predicted. Smart antenna technology greatly enhances 802.11n by being able to mitigate interference...picking the best TX/RX signal path for multiple simultaneous signals.

This turns Wi-Fi from a toy into a useful tool.

Comments

Post a comment

Comments are moderated, and will not appear on this weblog until the author has approved them.

If you have a TypeKey or TypePad account, please Sign In.