Skip Navigation

Wi-Fi causes a 30% drop in broadband speeds
Thursday 10 March 2011 12:15:54 by John Hunt

A study conducted by Epitiro, a company which specialises in benchmarking quality of broadband and Internet connections, has shown that on average, broadband customers see an average speed decrease of around 30% when using Wi-Fi to access the Internet.

The company recruited over 14,000 consumers in the USA, UK, Italy and Spain to test broadband performance when connecting over a wireless and wired connection to the broadband router. The average speed drop was 31.6%, with the UK faring best with a 26.5% drop whilst Italy came in worth with a 40% drop in speeds. The study did not however look to understand the difference between countries.

UK average download speeds, wireless vs wired
UK average download speeds, wireless vs wired. Source: Epitiro.
Click image to enlarge

Whilst speeds might be lower, the testing showed that this doesn't necessarily affect the quality of experience (QoE) when using the connection. Whilst throughput decreases, this isn't necessarily notable when browsing the web or using e-mails, but can be if downloading files. Equally, wireless connections saw an increase in latency and jitter (a variance in latency) when compared with wired connections, but this only affects specific applications such as online gaming or Voice over IP telephone calls using services like Skype.

"Our data shows that connectivity over Wi-Fi degrades broadband performance considerably in typical circumstances. Consumers who are experiencing performance issues with Wi-Fi should take steps to improve their home environment or connect directly via wired ethernet."

JP Curley, (CTO) Epitiro

Of those surveyed, 56% were connecting wirelessly whilst 44% used a wired connection. Unfortunately, wireless connectivity is on the rise, which means more people may start to see the impact of saturation on their network. As the number of wireless networks and the number of computers and mobile phones using wireless connectivity increases, there becomes less bandwidth available to use for each device. A quick check here and over 25 wireless networks are showing up as available.

Newer devices can be configured to automatically detect when congestion is seen and switch to using a different wireless channel to avoid problems which can help, but it is only a matter of time before the air-waves are completely full, and these problems become worse.


Posted by otester over 6 years ago
Only if your network is ****.
Posted by andrew (Favicon staff member) over 6 years ago
You can have a great wireless network, but still suffer the effects of signal degredation due to the physics of attenuation by material between you and the wireless antenna.

The research is really just highlighting to people that wireless performance is another factor to consider.
Posted by commandergc over 6 years ago
Talk about stating the bleeding obvious. WiFi does not have the through-put to use some of the faster connections! Well the consumer grade kit anyway.
Posted by mhc over 6 years ago
Eptirio's statement is WRONG.

WiFi does not degrade broadband performance what so ever.

It is true that the customer does not see the full performance of the Broadband connection when using WiFi but that not because the connection has been degraded but because the WiFi cannot cope.
Posted by MCM999 over 6 years ago
This report is based on a false premise. The connection sync rate negotiated by the modem with the source is in virtually all cases independent of the means by with the user connects to the modem. All this report confirms is that using a (possibly poorly configured) wi-fi connection is likely to be the rate determining step in the connection rather than the line sync rate.
Posted by Davidhs over 6 years ago
I agree with MCM999 that this report is confusing.
Four distinct issues here:
1. Modem sync rate. This depends on distance, line quality & IP profile set at the exchange. This has nothing to do with wireless. I assume Eptiro kept this variable out of their comparison.
2. Quality of wireless hw, sw & setup in router & PC. Most people may use wireless routers, I don't.
3. Distance & obstructions (walls, floors) between PC & wireless source
4. Local interference & bandwidth competition.

It is no surprise that items 2, 3 & 4 degrade performance compared with a wired connection.
Posted by otester over 6 years ago

When setting up a network you would take that into account ie: positioning of equipment.
Posted by Davidhs over 6 years ago
... if physical reality so permits! Not always the case.
Posted by otester over 6 years ago

Suppose not if you live in a igloo.
Posted by andrew (Favicon staff member) over 6 years ago
otester you may understand the nuances of positioning wireless kit. Does the average person on the high street?

The majority of people if they plug it in all using quick setup instructions and it works at all they will leave it set like that.
Posted by WelshBluebird over 6 years ago
I'd like to see the set up of a lot of these people. I'd be willing to be they live in older houses with thicker walls, possibly have cordless telephones (or other equiptment that used the 2.4Ghz band).
Posted by damien001 over 6 years ago
wifi will cause some delays even with the best kits, the overheads on wifi are more than on a full duplex wired lan. Also combined with that you got encryptions and other delays which you don't have on wired lan. Also the more wireless devices you have running on any one network the worse it gets
Posted by TerFar over 6 years ago
What a complete load of twaddle. I don't think that anyone should consider using Epitiro for any research.

My office is down the end of a long garden. My router (Linksys WAG320N)is located in a bedroom cupboard.

In my office I use an Edimax Wireless N USB dongle connected to a tiny external YAGI and get the full 11.1 MB/s that wired users get in the house.

What is the purpose of Epitiro's report? Did they actually have some goal when they set out to produce the report?
Posted by cpritch007 over 6 years ago
A chain is only as strong as its weakest link.
Posted by alexdow over 6 years ago
WiFi is also affected by many other, almost "invisible" or overlooked factors.

Your body as you use the keyboard, another person or pet walking past, precipitation etc. Vehicles and the like.
Posted by Oddball over 6 years ago
My wifi is still faster than my actual broadband connection. 11 Megabytes upstairs using wireless N vs 1.2MB download rate on the ADSL2+ modem. There is quite a way to go before most peoples broadband outpaces their wifi speed. Even on 54g.
Posted by toplessfly over 6 years ago
I would like to remind people that this research did not compare speeds on the same lines but between lines that used wifi and those that used only hardwire. I myself have a download speed 0f 9mbps a fraction of my wireless nominal speed 54mbps.
Anyway the speed of my line seems to be unaffected noticeably by the wireless connection that i use the vast majority of the time.
Posted by m101dream over 6 years ago
I assume all the people who disagree with these findings are all connected through wireless typical of them .
Posted by otester over 6 years ago

Ignorance isn't free.
You must be logged in to post comments. Click here to login.