Along with Netgear, D-Link is one of the first out with a product that "complies" with a draft of the IEEE's 802.11n fast WiFi specification. It promises a speed that is faster than ordinary 802.11g. We found that it does - up to a point.

When vendors build products to meet a draft standard, one expects that they will not work fully with other products based on the same draft, and users are normally advised to use a card and router from the same vendor.

Some early results have also claimed that in "real" situations, Draft N products can perform worse than plain 802.11g, especially as distance increases.

We compared D-Link's Rangebooster with Netgear's Draft N product in a terraced house where other wireless networks are visible, run by the neighbours.

In this situation, the tests we did could not be rigorous or repeated. However, we believe the experience would be fairly typical for a small office. Unless wireless tests are done in a dedicated facility, or use specialist equipment like that of Azimuth, your mileage may vary.

In most situations we found that Draft N gave some benefit over .11g, but it gave no benefit at all when the D-Link router was used with a different Draft N card. We didn’t have the capacity to really measure the full speed, but noticed it degrading over distance.

The D-Link router's performance was very variable even with its own D-Link card. To our surprise, in some cases, the D-Link router actually gave better results when linked to a Netgear Draft N router built from a different chipset (though not as good as the Netgear card talking to its own router).

Design and installation
D-Link has re-designed its routers: the Rangebooster is a black box with sporty-looking silver trim.

It has three adjustable antennae at the back, which D-Link feels is an advantage over Netgear's Draft N product (reviewed here) which hides them inside. The antennae are smaller and possibly flimsier than the spade-shaped ones on D-Link's previous MIMO products such as the G624M router (reviewed here).

The product also uses a small 5V power supply, not the 12V bricks most other wireless routers seem to need. Perhaps MIMO radios really do operate more efficiently.

It's clear that this is an early version of the product: the disks have paper labels stuck on them, and we were sent a new release of the driver for the card by e-mail. The card doesn't check automatically for driver updates - but this will be up and running soon, we are told.

It is possible, therefore, that any problems we experienced will be absent from the final product.

The device plugs into one of the LAN ports of our ADSL router. The installation routine works well enough on the router, and the management has all the features we looked for, including DHCP.

But the card isn't quite so easy. By selecting the Windows Zero Configuration option, we make the whole system hang. The uninstall doesn't work, so we have to use Windows XP's System Restore, to get it going.

Once installed, the card software is good enough, if less informative than the Netgear equivalent (there's no trace for the traffic level). The device includes WPA security but, as with the Netgear, no WEP.

More usefully for business users, the D-Link box supports WPA Enterprise, for those running an 802.1x authentication server. It also includes remote administration by Port 8080.

Disappointing performance
Now in actual use in our house, we found the D-Link router performed less well than the Netgear box, particularly when other vendors' cards were used. With its own D-Link card, it flagged up a connect speed of 243Mbit/s sitting next to the router, and transferred data a bit slower than the Netgear pairing.

In the next room, the symbol rate dropped to 121 Mbit/s and the actual throughput was much lower. In the room upstairs, where ordinary 802.11g does not reach, the card did connect, at an 81 Mbit/s symbol, but the throughput was even lower, with a 1.4 Gbyte file taking around 20 minutes. I make that about 1 Mbit/s.

We tried the Netgear card with the D-Link router, and got a marked reduction in performance. This pairing only made about 4 Mbit/s with the laptop on the same desk as the router, however, this speed remained available even upstairs.

We wanted to know whether this was actually better than g – some have claimed that Draft N is actually worse then 802.11g - so we set the router to 802.11g mode and found that, in multi-vendor settings, it was.

The Netgear card was now - in this 'g-mode - able to get data through the D-Link router faster, although we saw some odd behaviour – the signal went on and off for periodic blocks of 30 seconds.

Overall, the performance of this product was disappointing in our setting. We would like to believe this is due to some radio peculiarities of our house; if not, we hope that full production models are an improvement.


This is a good example of why standards need to mature before delivery. The situations where this gives advantage over 802.11g are too rare, and the performance is too variable to recommend it wholeheartedly.