commentary Wireless solutions -- they aren't all the same.
While RMIT is busily rolling out wireless LAN infrastructure around its campuses, quite a few issues have come to light -- AC power, security, and notebook WLAN performance.
Let me explain the first rather confusing problem -- AC Power!
It's all well and good providing hot spots around the campus in places such as the cafeteria -- students can work and surf on their notebooks while having a snack -- but what happens when their notebook also gets the munchies for some AC power? In the past, cafÃƒÂ©s were not renowned for supplying an AC point to every table so power points are at a premium. There are a few dotted about for cleaning staff and their vacs, but that's about it. Oh, sorry, nearly forgot -- vending machines. And there is the rub: we are seeing students unplugging vending machines to power their notebooks and then when another student tries to use the machine a minor altercation erupts.
RMIT is currently exploring security avenues for its WLAN but in the feasibility study rollout, security was ensured through the use of VPNs. Sounds fine until a visitor from another uni drops by and tries to use the WLAN only to find they do not have the same VPN software.
|When it came to WLAN in some cases it was simply "bung the card in and hope for the best".|
We have on numerous occasions been asked by organisations -- and even vendors -- to test the WLAN performance of various integrated WLAN solutions and PC Card WLAN options. The results of our tests were often quite surprising.
For example, in testing performed last year we found in some cases the tiny antenna in PC Cards provided far greater range than some notebooks' integrated solutions -- which is surprising considering notebook vendors have the opportunity to develop a larger and more "hardware matched" solution than the simple PC Card.
The latest series of tests, performed by the Lab just a couple of weeks ago, was on four different Centrino WLAN notebooks for a client to determine which of the notebooks had the greatest range. The notebooks under test were the Acer TravelMate 8005LMi, Dell Latitude D800, HP-Compaq nx7010, and Toshiba Tecra (Model: PT831A-67CST) -- a small but reasonable sampling.
Note that Acer claims they have "optimised" the WLAN performance of this particular notebook, so we were pretty interested (and a tad cynical as we always are when vendors make grand claims) in the outcome of the test.
We ran four series of tests, the first outdoors with clear line of sight, the second indoors with clear line of site, the third indoors with minor obstructions, and the fourth indoors with major obstructions. In the latter two cases "minor obstructions" consisted of typical thin office walls and major obstructions were thick brick walls and a large steel elevator shaft between the access point and the notebooks under test.
In the outdoor test the Acer was clearly superior to the other three notebooks and it turned out we could not test the Acer to its limit as we ran into a boundary fence at 183 metres. Even there the Acer was still connecting and providing good usable throughput while the other notebooks had all given up by the 132-metre mark.
In all the indoor tests the Acer also managed the greatest distance, particularly in the tests with minor and major obstructions.
There is quite obviously a lesson to learn here, a little tweaking of the integrated WLAN can yield some significant performance gains.
Steve Turvey is Lab Manager of the RMIT IT Test Labs, and can be reached at stevet@ rmit.edu.au.
This article was first published in Technology & Business magazine.
Click here for subscription information.