Wow thats bad

I originally intended to do a piece for the blog every couple of weeks or so but as I have the latest test data I may as well put it up a bit early.

So I got to thinking, I still have my original 2Mbps ADSL circuits, which both have wireless routers. Perhaps one of these will have a better wireless AP I could use instead of the BT Home Hub. I have the routers all connected to a shared switch so that I can arrive on any wireless AP but then route to the Internet via the new 40Mbps circuit. This is done by manually setting IP credentials on the laptop, in particular the default gateway and DNS IP addresses. So by changing the IP address for the default gateway I can push my Internet traffic via whichever DSL or fiber circuit I like independent of which wireless network I’m connecting to.  So should just be a case of repeat the original tests with the laptops connected to the AP I want to check, but default gateway always set to the 40Mbps link.

Diagram showing router interconnectivity

I have a 2Mbps ADSL from Sky (its free and good so why wouldn’t you) and a 8Mbps BT Internet ADSL which can just about manage 1.5Mbps - on a good day (distance issues). I’m intending to cancel the BT one once I’m happy the 40Mbps Infinity is stable and reliable. I’ll keep the Sky as its practically zero cost, reliable and will make a good backup.

I’ve already changed out the DSL router on the BT connection from the old style home hub to a Netgear DG834PN RangeMax that I bought in the hope it would improve the DSL connection speed – it didn’t. As it’s a newer router and has a better firewall (imo) I kept the Netgear and it’s been giving good service for the last 3 years or so. The Sky router is the standard one they shipped with the service – which I think is a rebadged Netgear as well.

Netgear Identify Parade: (left to right) Netgear WG602v4 wireless AP, Netgear G5608v3 8 Port GigE switch, Sky Broadband Router (looks like a Netgear DG834GT?), Netgear DG834PN RangeMax BT ADSL.

So I started testing the Sky DSL router wireless with high hopes of improving my wireless performance – not a chance. As soon as I moved out of the room the performance fell off a cliff – a very big cliff at that. In several instance connection was repeatedly lost during testing, preventing any result of the speed test.  

The following table shows the results for the Sky routers wireless.

OK so now I’m thinking perhaps my Sky router is seriously past its best so moved on to the Netgear DG834PN connected to my BT ADSL. This was definitely better than the Sky router as I would have anticipated,  but roughly half as good as the Home Hub and still some dead spots. As with the Sky router, wireless performance plummeted as soon as you moved outside of the same room.

Table for Netgear router “RangeMax” wireless results

This is what the blurb said that persuaded me to buy the Netgear:

 “With RangeMax dynamic connectivity there is no need to spend extra money on range extenders, repeaters and external antennas. There is also no need to throw away your older wireless devices; the DG834PN wireless modem router is compatible with 802.11b, 802.11g and SuperG™ products and will boost their range and speed. Coverage is maximised when the DG834PN is used in conjunction with NETGEAR’s RangeMax™ PC cards and adapters.”

Well it’s definitely doing something better than the Sky router but its performance was still down compared to the Home Hub.

In summary, here’s the Home Hub results from the last post and the Sky results in a single table for comparison.

I’d include the Netgear as well but the table is too crowded, you get the idea anyway.

Conclusion

• My Sky routers wireless performance is plain awful compared to the Home Hub, its just worlds apart.
• The Netgear with “RangeMax” technology was able to give better speeds but struggled with reach almost identically to the Sky Router.
• All wireless routers are not created equal and you cant use ABG/N specification as a measure of how good they are. Spending cash on a router doesn’t make it better than a free one.
• The technique for using different wireless vs gateways works well and will allow me to swap between Sky and Infinity with a software switch. Also allows for different APs to be introduced if I can find one with better performance than the standard Infinity Home Hub.

Next

Support call to Sky to see if the performance I’m seeing is expected behavior and if they can offer anything better. If they do, I will test it and share the results.

Trawl of forums to see what the community has to say on the subject.

Check router product reviews for a new candidate for wireless networking tests. 

As I’m diss’ing people’s equipment here and they probably have big legal teams just waiting to sue bloggers I probably should get me one of those legal disclaimer things so here goes:

All data and information provided on this blog is for informational purposes only. Technical-stuff makes no representations as to accuracy, completeness, currentness, suitability, or validity of any information and will not be liable for any errors, omissions, or delays in this information or any losses, injuries, or damages arising from its display or use. All information is provided on an as-is basis and confers no rights.

This is a personal weblog. The opinions expressed here represent my own and not those of my employer or probably anyone else for that matter. I make no claims to expertise or subject matter specialism, its not unusual for me to be wrong either. You get the idea - I’m just doing stuff for my own amusement.

  

Wireless, how slow can it be?

I’ve just been hooked up to the BT Infinity 40Mbps broadband service after spending almost a decade stuck with 2Mbps ADSL. So while testing my new broadband on www.speedtest.net I found that I wasn’t getting the full 40Mbps.  Soon identified my wireless network to be the culprit, it struggled to get anywhere near as fast as the new Internet link. When I moved my laptop further away from the Wireless router performance seriously dropped to around half line speed. Conversely, when hooked up via LAN cable, the speed was about the 40Mbps I’d expected.

Thinking perhaps my laptop was past its best I set about checking the wireless performance of a selection of laptops, netbooks, I Phones – pretty much anything with a wireless capability really, from various points around the house. Well turns out my laptop is fine but wireless kind of sucks J

This table shows the performance I measured over my wireless network for the selection of devices and locations in my home.

You can see the max speed for my Internet line is 37.65 Mbps (down). Once you add wireless into the mix it gets nasty. Look at the I-phone 3GS – 15Mbps max and that’s when its 12” from the wireless router. 1^st gen Touch I can kind of forgive but 6Mbps is pretty slow by my new standards. The Thinkpad I thought might be getting old turns out to have by far the best wireless connection – who knew.

So it’s pretty clear to me that:

• Wireless is great for flexibility but sucks for raw performance
• As a rule of thumb I cant expect to get much more than 20Mb from the Internet if I connect via wireless and probably a hell of a lot less (ignoring same room, I’d use Ethernet in that instance anyway)
• I can get downloads at least twice as fast by using cable into the hub rather than wireless.

The Lab

Whilst this isn’t a laboratory environment by any stretch of the imagination I should explain my test “environment” aka my home. Firstly I picked 4 “distant” locations in the home at which to perform the tests. These where:

1.      Study, on my desk ,12 to 18 inches from the AP router

2.      Far end of the lounge, this is across a hallway from the Study, about 12m from the AP

3.      Bedroom 1 3m elevated and 12m from AP, above Lounge

4.      Bedroom 2 3m elevated and 10m back from AP

At each test location I used the local browser (mostly ie8) to access www.speedtest.net and performed a speedtest to the recommended test server (Milton Keynes – my home city). For the apple devices the speedtest app was used instead. All NICs where set for auto speed though clearly some aren’t capable of N rates. As the AP remains constant the only other major variable is the wireless capability on the relevant laptop.  No attempt was made to run latest drivers, this was purely a "suck it and see" test.

Next

I’m going to check out my other two wireless routers to see if they are any better than the BT Home Hub