Labs

Wireless-n routers

Thanks to the certification process, this is the first wireless routers Labs in which we've been able to test using a "standard" adapter rather than relying on manufacturer's own adapters.

To take advantage of this, and to ensure the routers worked well with other manufacturers' kit, we tested with Intel's near-ubiquitous 4965AGN laptop adapter, embedded in a Lenovo ThinkPad X300.

Aside from interoperability, though, there are a number of other factors that make testing wireless routers tricky. Interference from neighbouring networks can affect speed and stability, and there's a whole raft of other devices that can impinge upon performance.

Digital cordless phones, for instance, operate in the same 2.4GHz range as Wi-Fi, and microwave ovens can also play havoc with your connection. This means performance can vary wildly from fast in the morning to dead slow later on.

To make our tests as real-world and repeatable as possible we tested all the routers in domestic conditions - a small, three-bedroom suburban London terrace - and, to ensure a level playing field, took control readings from a draft-n 2.0 router not involved in the Labs (an EnGenius ESR-9710) before each router was tested. The final results were then adjusted depending on the relative performance of the control readings.

For each router, we performed a number of timed file transfers: we copied 128 files 1MB in size from another laptop connected to one of the router's LAN ports and repeated that three times; then we did the same with one 128MB file. Next, we watched a two-minute 1080p WMV video clip across the network to see if each router could maintain a steady stream of data.

We performed this set of tests in a variety of locations around the house to gauge performance at different distances. The first set was conducted in the same room at a distance of 2m; we then moved next door - 3m away with a brick wall in between; we tested in a bedroom, one floor up; and we did the same in the kitchen, separated from the router by a staircase and two brick walls.

Finally we took the laptop out into the garden and into an office 40m away, with a double-glazed window in a brick wall, light vegetation and a 19mm-thick wooden wall in the way.

All routers were tested at factory-standard settings with WPA2 and AES encryption turned on and the latest firmware version applied. We also ensured that each router was using a channel that other routers in the area were not using.