4. The Test

When I tested the WSB24, I used a combination of NetStumbler and Ixia's Chariot to measure the effect of its boost. This time around I still used Chariot, but also - thanks to the nice folks at AirMagnet - was able to use AirMagnet Surveyor [reviewed here] to provide a very helpful graphic display of the before and after.

Figures 3 through 10 (click on any to get a larger view) show Surveyor plots of Signal strength (left column) and Speed (right column) measured by an active survey of the main level of my home. A Surveyor Active survey is done by having the wireless card - in this case an Atheros reference design dual-band card - in the notebook running Surveyor, actually associated with a target AP - in this case a Linksys WAP54G version 2.

There are four sets of plots representing no BA24j's used, one used on the WAP54G left antenna position only, right antenna position only, and used in both antenna positions simultaneously.

In each of the Signal Strength plots, I had Surveyor show a balloon indicating Signal, Noise, S/N, Speed, Retry Rate and Packets Lost at a particular spot. I also limited the Signal Strength plots to show Signal above -60dBm only. I picked this value after experimentation and it seemed to best illustrate the BA24j's boost effect within the area that I surveyed.

Figure 3: Signal - No Boost	Figure 4: Speed - No Boost
Figure 5: Signal - Left Boost	Figure 6: Speed - Left Boost
Figure 7: Signal - Right Boost	Figure 8: Speed - Right Boost
Figure 9: Signal - Both Boost	Figure 10: Speed - Both Boost

Even without opening larger views of the plots, you should be able to see that using only one BA24j in either position widened the -60dBm signal area, and that the widest area was achieved with using two Boosters. The signal readings taken at the "balloon" locations show -63dBm with no boost and -55dBm, -51dBm and -51dBm levels for left only, right only and both boost conditions - a measured gain of 8 - 12 dB.

The set of Speed plots show that the main effect of using the BA24j was to reduce the area where the AP and STA were communicating at a 36Mbps rate. It's interesting to note that for the combination of AP and AirMagnet card that I used, boosting the signal didn't maximize the use of the 54Mbps connect rate, but just seemed to shift it around and break it into pieces.

I also moved the WAP54G to the room at the upper left corner of the floorplan above and ran Chariot throughput tests with my notebook loaded with a Linksys WPC54G card located in the room at the upper right corner of the floorplan. I wouldn't really call the results conclusive, since the variation in runs could just have well been produced by the WPC54G shifting connect speed ranges (I didn't force transmit rate either on the AP or STA). But it did seem that the Booster in the left antenna position produced the most noticeable throughput improvement.