Epitiro - is that welsh for 'cowboy'?
Mine's the one with the Ethernet cable in the pocket.
You'd never guess this. Connect to the internet in your home over a wireless connection and you won't get the same data throughput as you do if you connect using Ethernet. Amazing, and thank providence we have Cardiff-based network testing company Epitiro to tell us these things. It bases this staggering conclusion after …
Epitiro - is that welsh for 'cowboy'?
Mine's the one with the Ethernet cable in the pocket.
The pope is indeed catholic
Bears and woods.
... when your 100Mb Ethernet connection is choking the internet! :(
So Speed... is that bandwdith or latency... and does drastic latancy affect measurements of bandwidth?
because most people I know have wifi that connects at 3x thier isp bandwidth so loosing 30% at the wifi still leaves ample capacity to max the ISP connection.
The people I rent from use 802.11 N to connect to their ADSL (7MB down /750k up) but don't get anywhere near the speed I do with my more direct connection.
Remember that unlike WIFI, Ethernet and ADSL are full duplex(can send and receive at the same time) and that will make a large difference in both throughput and latency.
"All this is, of course, well know to geeks like you and me. To be fair, Epitiro has less technically savvy punters in mind"
Well, yes, accepted. But they didn't have to do a huge survey to find this out, did they? They could have just mentioned it (because, presumably being geeks themselves, they knew this anyway).
What next? Eating more greens is healthy for you? Oh, hang on.. I think we've done that one...
They might know that WiFi is slower than ADSL.
However, /qualifying/ just how slow WiFi is in real-life is a different matter, that is best done on punters equipment, and during normal usage hours (when the neighbours WiFi is likely to be at its most conflicting).
That's what you get for believing marketeering numbers. 54Mbit? I'll take 100Mbit FDX ethernet, thanks. 300Mbit? I'll still take that ethernet, thanks. Or, given today's prices, gigabit.
Pick a different channel? Amazingly, that likely isn't going to help. The 2.4GHz band comes in 13 channels in Europe --only 11 in north america and some manufacturers "forget" to make the extra channels available here-- but only about three, maybe four, usable ones, due to overlap. Punters that think to be smart by picking another channel but don't know about the overlap are just causing interference for both adjacent previously usable channels.
Similarly, "802.11n" gets touted now like it's all you need to know. Not so. If you're still in the 2.4GHz band, all that does is eat so much bandwidth it fills up the entire band, choking up the rest. That is good marketeering because it causes more people to "upgrade" to 802.11n in the 2.4GHz band, pushing hardware sales. But it doesn't alleviate the problem much.
Still and all, wireless is useful, so if I get the choice, I pick the kit that also does 5GHz, but since that sometimes just isn't the case I pretty much have to run both. *sigh*
And Tony, don't forget to mention which bands are supported in your reviews. Thanks.
We always mention the bands an 802.11n adaptor supports as a matter of course now.
You will see either '2.4GHz 802.11n' or '2.4/5GHz 802.11n', as appropriate.
BTW, since 802.11n incorporates 802.11b, g and - in the 5GHz case - a, we don't mention the older specs too. In case you were wondering.
doesn't the 802.11 standard require that the fastest possible speed (Mbps) will be dictated by the slowest device connected to it? So an "n" network will only run at a maximum of (for example) 11Mbps if you have a "b" device connected - regardless of what standard the rest of the connected devices are.
And that's before allowing for interference from neighbours, steel beams, distance etc.
But you don't have to accept that. That is, g APs usually have an option to refuse to talk to b-only kit, and then you don't have to drop down. Though I don't know what happens if there's a "b" conversation on the same channel but with a different AP. Same with 'n' and b/g or a.
"n" is also supposed to be lots more resilient but I haven't really tried yet; no 'n' capable station equipment (yet) and the AP is a thomson so any results will be spelled "crap" anyway.
mp3's are classed as "large files"
Next they'll tell us that mobile broadband isn't as quick as wired broadband...
say it isn't so!
I live in France. In the country. In the middle of nowhere. I went out the other day to test my newish Android phone and speedtest.net (averaging three tests) says my phone, while I can see nothing but fields, provides 2200kbit/sec down and 960kbit/sec up. My wired internet, at the end of a 4.7km wire, is 2048/256. ;-)
My only question is why the phone (Motorola Defy) tops out its WiFi capabilities at around 900kbit? It just doesn't seem capable of running faster, even though it obviously can on 3G. Bizarre...
the real problem is that most people's method of solving congestion and interference is simply to crank up the power and fit high gain antennas in order to "shout the loudest".
the correct solution is for everyone to run the appropriate power sufficient to cover their property and no further. however, that requires everyone to do it, if 95% of people do it but the rest don't, the ones who crank up their power will adversely affect the others.
Or does this seem like the biggest waste of resources in the histroy of the earth, any 16 year old with an IT GCSE could have told you this
Most of them barely comprehend the IT lessons they have already, the ones that teach them how to use MS Office ( well Word and and Excel! ) and the end of term lesson is that "da net" is a giant thing called a network!
Unfortunately, this is not the case. When I was going through Uni, the instructors for the computer-related classes (programming for instance) barely had a cursory knowledge of the subject. In the case of one particular programming teacher, they spent the summer prior learning the language.... At least the hardware instructor understood hardware. Problem is, they knew crap about software and couldn't figure out why his corrupt Windows user profile (well, corrupt user registry in his profile) was causing his machine to do the whole log-in-but-immediately-get-logged-out issue..... Unfortunately, those most qualified to "pass on knowledge to the next generation" are working other jobs making boatloads more than the pittance given to teachers. Once in a great while you'll find a technical-genius gem of an instructor who loves to teach more than he loves a decent paycheck. Those are the ones that deserve awards and recognition.
/grammer nazi, since it's the only academia-like icon available....
This needs a protest group and I already have the catchy chant all worked out.
What do we want?
Faster hardcore prawno!
When do we want it...uurgghe ....Never mind!
I'll get me coat....it's the grubby mac over there
was a group meeting, would anyone come?
Was round ex's the other day, and she was complaining how bad her ISP was, and that they got a really terrible connection, and Virgin were rubbish etc.
Then, after a couple of minutes at the PC I pointed out the problem was her Wifi connection dropping out completely and probably not the ISP at all.
Didn't have time to check it, but I expect she is on a crowded channel. Which she had never even heard of being a problem.
Remember, there are reg readers, and normal people.
She is "normal"?
... why it didn't worked!
My understanding is that 5Ghz isn't so hot through solid objects like brick walls - anyone done any testing on this? Given I can see 20+ SSIDs from my London mid-terrace on a good day, all on 2.4Ghz, there's an argument for moving frequency - but not if I can't get through the walls...
If you're talking solid concrete walls then neither is going to be noticeably better/worse.
If its bricks then 5GHz should be better as the the bricks will contain some water content which absorbs more of the 2.4GHz signal.
What's worth noting is that you're permitted to transmit at 3-9dB greater ERP as free space losses @ 5GHz are higher than at 2.4GHz. Should make a nice difference in the average UK house - although it won't last as adoption rates increase.
Personally I'd just bite the bullet and wire up the rooms in your house with CAT5e or CAT6 (if you're decent at terminating wiring). Don't bother with the rooms where the wiring would be too exposed/messy - use a dual-band 2.4/5GHz AP there if you're starting from scratch.
This article should be put into an urban myth website.
How can you come up with reliable statistics on this? The issue is more to do with the environment that the technology is applied to as opposed to the technology itself.
WiFi is not the issue, it's people's homes, neighbours Wifi signals interfering whewther it's a terraced flat, or a large block of apartments with different levels.
...as we have a HomePlugAV (200Mbps) network around the house for anything which really needs the speed. Only my netbook and our smartphones use WiFi at home, and they don't generally need a cookin' connection...
Still, seems a bit of a "well, duh" story, but then it would to this audience, wouldn't it?
(mine's the coat with the asbestos lining, in case the radio hams read this and spot the HomePlug reference...)
As mentioned above, it is an environment problem.
Using wireless G at 54Meg, I get a throughput from my VDSL (50Mbps) net connection of around 22Mbps, so in this case yes, it is slower. With wired ethernet I can at off peak times get the full wack 50Mbps!. If I use wireless N I can also get the full wack 50Mbps.
I have changed the channel on the Wifi, as I live in a building in the 'innenhof', a building in the middle of a square inside an outer block of apartments, so many other AP's visible.
It is fantastic broadband though, off peak I have downloaded Ubuntu live CD's in 2 minutes.
.... there was an advice label on the boxes of the last two wireless routers I have bought stating that wireless connection could be slower than a direct cabled connection.
I'd love to know who approved the funding for that research. What a mug!
Been there, seen it, done it, thrown the kit away.
Nice empty 5Ghz band around here and there's a reason for that. Something stomps across the entire band with big spiky boots on a regular basis, knocking out all the connections. For added fun, while various manufacturer's Wireless-N kit seems to play nice in 2.4Ghz, interoperation is lucky lotto time once you get into the wonderful world of 5Ghz. I found that out the hard way.
At least in the congested (and it is here, as 1-4 are unusable for a similar and unknown reason) 2.4Ghz band, I can get a solid signal with a consistant and measurable 140Mb/s throughput. This also works between floors of the house, something else that the 5Ghz kit couldn't do to save its life.
Have you forgotten the first rule of 5GHz club?
I live in the US. Even with a 30% drop in the connection speed of my g wireless I'm still almost twice as fast as the fastest ISP connection I'm willing to pay for, $40/month.
There is a little app on my phone that can display all the WiFi channels and their signal strengths. Sometimes when I am out in town having a meal, I fire it up to see what is around. I'll tell you what is around. There is one smart guy with the world to himself on channel 1. Then there is everybody else, bang smack in the middle of the band, all on top of each other. I'd guess there are seven to nine APs, but it is hard to read.
The app is "Wifi Analyser", I'd recommend anybody with an Android phone to try it to find a nice quiet channel you can call your own, instead of you router's default setting of "let's hog the middle lane".
That's the bottom line. I run Ethernet or Ethernet over mains. I threw the wireless stuff out long since Ethernet must be at least 50% faster and sustained. I have a wireless laptop but even that gets plugged into wired when I want to download/update software etc. My sky router won't transmit enough wireless to even connect between floors. I did have a conversation with a lady from sky about channels which was one of those wasted moments I'll live to regret.
wgetis broken and should DIE, dev tells Microsoft