Gallery

Security without penalty

Posted on 30 Mar 2009 at 16:17

Paul Ockenden reduces the strain on his netbooks, continues his Wi-Fi investigations, and reveals the perils of using a mobile as a modem.

A second problem is that being a computer itself, Gatekeeper draws power and so will sap some of your battery time. It consumes only 2W at max, but that's fairly significant on a low-power netbook - on a par with the current draw of a Wi-Fi card. Overall, though, it gets a big thumbs-up.

New adventures in Wi-Fi

My investigation into dual-band 802.11n kit is still ongoing, and I have a few more interim observations to share this month. The first concerns MIMO (multiple input, multiple output) devices in general, which includes 802.11n - whether dual-band or 2.4GHz only - and also some speed-boosted 802.11g devices. Essentially, it means any router or access point that has more than one aerial. You might have noticed that when 802.11n kit first arrived it was normal to see three antennae poking up from the back of routers, ADSL modems and access points, but wander around a PC superstore these days and you'll see that most of the kit now sports only two. So is the dropping of that third aerial simply a cost-cutting measure?

To some extent perhaps, but there's more to it - if you Google for terms such as MIMO, spatial multiplexing and diversity coding, you'll find lots of headache-inducing maths concerning the way multiple antenna systems work: in essence, the more aerials you have at both ends the better, but if you have three, chances are only two are being used for transmitting and the third is reserved for reception (generally known as a 2T/3R configuration). There are a few 3R/3T access points out there, but they tend to be very expensive, and I don't know of any that are targeted at home or SME buyers. Having said that, there are several 3T/3R mini-PCI ExpressCards available, so it's easy to upgrade most laptops to a full three-radio configuration. Even if one antenna is used only for receiving, it's still obvious that a three-aerial router used with a three-aerial laptop will in general give better performance (especially in suboptimal locations) than two-plus-two will.

The problem is that most laptop vendors have been shipping their 802.11n-enabled machines with two antenna (2T/2R) Wi-Fi cards installed, so you'll get a 2T/3R to 2T/2R configuration if you buy a three-aerial router, or a 2T/2R to 2T/2R setup with one of the more recent twin-aerial devices. Since I struggled with the mathematics (my Physics degree is a dim and distant memory), and I'm more interested in real-world performance, I chose to adopt an empirical approach by performing some tests around my house and office using various combinations of client adapters and access point/routers. I discovered that although 2T/3R at both ends (router and laptop) was always significantly better than a dual 2T/2R configuration, in practice having 2T/3R at one end and 2T/2R at the other was only marginally better than dual 2T/2R, and in many situations the performance was pretty much the same. Given which, it's hardly surprising that a router manufacturer might cut costs by dropping the third receiving antenna from consumer-aimed products.

Although it may not be a surprise it's certainly a pity, because it's likely to discourage laptop vendors from including three-antenna radio subsystems in their products, and so unless we shop very carefully, we consumers are going to be denied the full benefits that 802.11n could bring. In fact, the full specification of the 802.11n draft standard allows for four-aerial 4T/4R configurations, but I'm not aware of any chipsets that support anything beyond 3T/3R.