802.11a vs. 11g in Homes

In my previous tutorial,
I discussed how to setup a home wireless LAN. Let’s take a closer look at
the range and performance that you can expect from a single access point in
a home environment.

For this testing, I setup a typical home variety dual-band wireless
LAN router in the center of the first floor of a 2,700 square feet two story
house with a finished basement. The router implements 802.11a and 802.11g, and
channels were set to 6 and 52, respectively. Transmit power was set to maximum
for both. Standard omni-directional antennas and default router configurations
were in use. In order to measure signal strength, noise level, signal-to-noise
ratio (SNR), packet retries, data rate, and throughput, I made use of Airmagnet Trio.

In order to achieve best results, I made sure that RF interference
was kept to a minimum. There were no microwave ovens or cordless phones in use,
and Airmagnet indicated no other access points in the area.

Best-Case Results

I first took measurements within a couple feet of the router. This of course
is where the signals levels are highest, where you’ll get optimum performance.
The following are readings at this point:

Even though both 802.11a and 802.11g were operating at 48Mbps at this point,
the throughput of 802.11a was higher at 2478 packets-per-second (pps). The reason
for this is that the noise level (-86dBm) in the 2.4GHz band (802.11g) was considerably
higher, which resulted in lower signal to noise ratio (SNR) and corresponding
higher packet retries. 3% of the 802.11g packets sent had to be retransmitted.

Worst-Case Results

I took many other measurements from different places around the home, such
as in the far reaches of the basement and within bedrooms upstairs. The following
shows the poorest readings that were found, which was in the corner of the master
bedroom on the second floor:

The signal strengths for both 802.11a and 802.11g are plenty high enough to
maintain continuous associations. The data rate of 802.11a, however, shifted
down to 36Mbps due to relatively low signal strength. The lower noise level
in the 5GHz band, though, kept 802.11a packet retries very low (0%) as compared
to the higher retry value (8%) found with 802.11g. This led to 802.11a having
higher throughput at 1877pps, despite the fact that the 802.11a data rate had
downshifted to a lower rate than 802.11g.

What does all this mean?

The most significant conclusion that I draw from this testing is that a single
802.11a or 802.11g router (or access point) is enough to provide good performance
throughout an entire home. This assumes, however, that you can install the router
in a central location.

The router may need to be next to a wall on one side of the house, nevertheless,
if that’s where you connect to the broadband service. This could make signal
levels on adjacent corners of the house too weak for some applications. The
presence of RF interference from operating microwave ovens
and cordless
phones will also causes packet retries to increase, which lowers throughput.

Something else that these results indicate is that 802.11a offers higher throughput
in all parts of the home. At best, throughput of 802.11a was 16% better, which
slightly increases file transfer speed. Keep in mind, through, that 802.11a
will not experience the dramatic drop in throughput that 802.11g does when RF
interference is present.

Which one is best for Homes?

802.11g makes most sense for use in homes. This standard supports laptops with
802.11b radio cards, and 802.11g wireless clients will have good performance
and be able to interoperate with the prevalent 802.11b/g wireless networks found
in enterprises and public hotspots. If you’re a heavy user of 2.4GHz cordless
phones, though, strongly consider using 802.11a.

Jim Geier provides independent consulting services to companies developing
and deploying wireless network solutions. He is the author of the book, Wireless
LANs and offers training focusing on wireless LANs.