Sponsored By

The Dark RoomThe Dark Room

I spent most of the weekend locked in one – trying to answer an important question

September 27, 2006

2 Min Read
The Dark Room

5:05 PM -- I spend a lot of time in dark (well, poorly lit, anyway) rooms running benchmarks on wireless LANs. Mostly I’m doing the usual, comparing the performance of different implementations, trying to find out how well a given product works, often in comparison to one or more of its brethren. If you’re a nerd like I am, this is, believe it or not, fun. And it’s also important both to vendors and those who are purchasing the equipment for production applications. We’ve seen huge variances in performance even from products based on the same chipset, let alone those with dissimilar architectures. By the way, I will shortly publish our recommendations for how to benchmark relatively small-scale WLAN configurations, and I will have this document for you here.

But this begs the question: How does one benchmark really large WLAN configurations? It’s impractical to install a big system just to test and evaluate it, so a preferred approach, used in the design of everything from aircraft to nuclear weapons, might be to simulate the system, extrapolating to the real thing. This is the concept I set out to test.

So, what I did was borrow a VeriWave Inc. WaveTest system which I used to establish the simulation environment. I also got some time in a large, walk-in “RF cage,” which is formally known as a Faraday Cage, where I ran corresponding but not identical real-world tests. The Faraday Cage is made out of copper screen, and no RF energy gets in or out of it, allowing experiments to be run even with nearby WLAN and other systems in operation. I performed tests on enterprise-class systems from both Aruba Networks Inc. (Nasdaq: ARUN) and Meru Networks Inc. , also borrowed (consultants and analysts are notoriously cheap and/or broke, it seems). I then compared the two sets of results and assumed success would be indicated by a strong correlation in results between the two environments.

I’m still evaluating the mountain of data generated from the weekend’s worth of time spent on this project, but I’ll write this up over the next few weeks and post it for you here. My preliminary conclusion is that we’re going to get a lot of mileage out of the simulation route, and I’m excited about the potential here.

— Craig Mathias is Principal Analyst at the Farpoint Group , an advisory firm specializing in wireless communications and mobile computing. Special to Unstrung

Subscribe and receive the latest news from the industry.
Join 62,000+ members. Yes it's completely free.

You May Also Like