It's networking Jim, but not as we know it

· by Steve · Read in about 5 min · (996 Words)

As regular readers will already know, I recently made my choice between 360 and PS3, in favour of the 360 (on the basis of the view that right here and now, it best matches the gaming experience I’m personally looking for). After that it really just came down to practicalities, and one of the things about the 360 is that it has no built-in wireless. An odd decision in hindsight, I can only assume that it was decided early on when the machine was being designed, and that they couldn’t add it in later without having major space or shielding issues - certainly the cost of a wireless chipset is low enough now not to be an issue, so I assume it’s a design problem related to fitting wireless in with the other components.

The extra £50 they want for 360 wireless bridge is daylight robbery so I went looking for alternatives, since my router is upstairs in my study/home office. One of the things I want to try when I get the 360 is streaming media from my Linux server upstairs via something like uShare, and it also occurred to me that as time goes on, the chances of me wanting more network connections in my TV cabinet are only going to increase, and much of that will undoubtedly start being heavier multimedia streams, for which wireless can be a bit slow and unreliable. Let’s face it, the 802.11g specs might say 54Mbps, but you’re unlikely to get that unless you place the aerials 1cm apart. In more realistic scenarios you’re probably getting about 20-30Mbps, and perhaps even less if you have a large distance to the access point with solid walls / floors in between, or you have many wireless devices fighting over the signal.

So, I decided that for a fixed cabinet (as opposed to something like a laptop), powerline ethernet made most sense - this is where you route the network connection through your home’s regular electrical wiring using spare frequency bands. It still suffers from interference of course (show me a signal that doesn’t), but there are fewer variables involved and the maximum rates are faster. I chose the Netgear HDX101 which has a theoretical transfer rate of 200Mbps, although I knew I’d never get that (I only have a 100Mbps hub currently anyway). There are standards called ‘Homeplug’ which cover this powerline networking, of various speeds, the latest being Homeplug AV which is 200Mbs. As it happens, this Netgear product supports a proprietary format instead, but I chose it because it’s reportedly faster than the current Homeplug AV kit, and I managed to get it on a special offer which made it a lot cheaper than the standards-based ones which are newer. I only need point-to-point right now anyway, I can always replace them if compatibility becomes an issue later. There is also 85Mbps kit available, but it wasn’t much cheaper than the special offer I got, and I figured I might as well get something with a higher potential rate so that once interference was taken into account I got the best speed I could. Reviews of the HDX101 were extremely variable - some people’s testing suggested bad connection rates, some sang it’s praises. Given other 200Mbps hardware was twice the price of the offer, I figured I’d take a chance and try it.

Installation was a snap - plug both in, switch them on, wait 30 seconds. Done. 😀I installed the software on my PC just so I could see what the reported speed was, and it started off at  a pretty respectable 65-70Mbps. However, on checking again a couple of hours later, that had risen to just over 100Mbs, obviously touching the maximum bandwith I can make use of with my 100Mbps hub anyway. I’m not entirely sure why it increased, since the number of electrical items switched on in the house (a potential source of interference) was identical. It’s not a statistical blip either, I’ve tried it multiple times and if anything it’s still getting quicker (last test was 105-107Mbps). The Netgear is supposed to have some kind of proprietary optimisation routines (part of why it’s not Homeplug AV compliant) so maybe it just takes a little while to figure out the best frequencies to use given the particular setup of our house. It certainly performs far better than the negative reviews have suggested, and agrees with or even exceeds the positive reviews I’ve read. Who knows what the variables are - perhaps it works better on UK wiring? Our house’s wiring is not exactly that new, but UK wiring has to be able to take more of a pounding than US wiring - because our standards were set in the optimisic 60’s when they thought everyone would run their house’s heating on nuclear powered electricity, and might want to, you know, run an arc welder on a regular plug socket for example) - maybe that’s a factor.

So, that went even better than I expected - I now have 4-5 times the bandwidth in my TV cabinet than wireless would have provided (or more), and the QoS should be better too, which is important for video. Plus, I have the ability to extend that fast network to other items in the TV cabinet later on without all the cross-talk and interference you’d get by cramming multiple wireless devices in there. Splendid! 😀

Update: having read more about this, it appears you may actually get poorer results if your wiring is pretty new (and particularly in the US), because of all the separate circuits with breakers they have to put in now. Really old wiring is also bad because of the corrosion and potentially sub-par materials, the sweet spot is probably in the middle, which I’m guessing our house is. I expect this is why so many review sites thought this equipment didn’t work well whilst I saw many other individual people reporting good results, as I have.