Wireless Performance

NOTES:
• "Signal Quality" readings were not available
• WiFiSec mode enabled and WEP disabled
• Testing was done with an ORiNOCO 802.11b Gold client card in a WinXP Home Dell Inspiron 4100 laptop

I started out trying to use a NETGEAR WAG511 a/b/g card [reviewed here] for my testing, since it has a good 11b radio and I wanted to give the TZW a fair shot at turning in good performance marks. However, after futzing around with it and other draft-11g based cards for a few hours I found that I couldn't get any card to go faster than about 1Mbps or so. So I gave up and decided to use an ORiNOCO 802.11g Gold card instead. I later read the following in the release notes for version 1.0.0.1 of the TZW's firmware:

Some wireless cards, notably Netgear and Linksys 54G cards, have demonstrated compatibility issues when connecting to other manufacturer's devices. Compatibility issues are further complicated by the use of WEP, or by increasing the beaconing interval beyond its default value. It is strongly recommended that you upgrade to the latest firmware and device drivers available from your wireless client card's manufacturer, particularly if you are using Netgear or Linksys 54G equipment.

I asked Sonicwall how they were sure that the problem lay with the draft-11g cards and not with the TZW's radio and basically got a "draft-11g new and unstable, PRISM stable and mature" (I'm paraphrasing) response. But no matter what the cause, I'd have to say that I agree with Sonicwall's warning and advise not planning on using any 802.11g cards with the TZW until after 11g is ratified and you update your card's firmware with released 11g code!

With my test client chosen, I then decided to do my four location throughput test in the WiFiSec mode and without WEP enabled, since I figured most users would be running in that mode. Figure 11 shows the results.

Figure 11: Four Condition Throughput test - WiFiSec mode
(click on the image for a full-sized view)

As you can see, the TZW had consistent throughput in all of my test locations, but with 3Mbps average throughput - significantly lower than the best case throughput of good 802.11b products.

I attribute the consistency to the dual 5dBi antennas and 200mW radio which together produce a stronger signal than the 2.2dBi dipoles and 30mW radio that come with many consumer APs.

The lower throughput, however, came from the use of WiFiSec. Since the TZW has two potential throughput stealers in its WEP and WiFiSec encryption options, I ran some tests to quantify the effects of each.

Figure 12: Security option throughput comparison
(click on the image for a full-sized view)

Figure 12 shows four Chariot "Condition 1" (close range) runs with:

You can see that the best case wireless throughput is around 4.5Mbps, with WEP taking an almost 20% bite and knocking throughput down to about 3.7Mbps. Choosing the more secure WiFiSec, however, lowers throughput by about 1/3 (33%). Fortunately, the effects of WEP and WiFiSec aren't additive, since the last run made with both enabled shows the same 33% reduction as with WiFiSec alone.

I found this result surprising, especially considering that my Ethernet-based VPN testing showed about 7Mbps available through a wired tunnel. But there are apparently other factors that come into play with the wireless WiFiSec that result in the reduction that I measured.

So although I give the TZW excellent remarks on throughput vs. range performance, I have to say I'm disappointed with the significant WiFiSec throughput hit. 802.11b's best-case 5Mbps-or-so throughput is marginal enough for busy wireless LANs, and business users may find the lowered WiFiSec bandwidth insufficient for their needs.

802.11b Wireless Performance Test Results

See details of how we test.