Join or sign in Register for your free TechRepublic membership or if you are already a member, sign in using your preferred method below. Use Your Email Use FacebookUse Linkedin

WiFi Effect on Download Speed

Locked

I have searched high and wide for this subject in forums and white papers of all types and I have yet to find a succinct explanation of what is going on here. Do you know of any white papers or threads that would explain this?

These are my results when testing some equipment:
ISP Bandwidth package = 30Mbps
Wired to Modem = 30Mbps D/L Speed
Wired to Router = 29.9Mbps D/L Speed
Wireless via 802.11g to 802.11g client = 15Mbps D/L Speed
Wireless via 802.11n to 802.11n client = 26Mbps D/L Speed

Is this sort of loss normal when using WiFi? Is there a scientific explanation of what to expect in terms of % of D/L speed lost for each type of 802.11 connection? In other words if you have a transfer speed of 54Mbps, do you normally lose 50% of your Download Speed compared to wired?

Thanks!

Topic

Clarifications

In reply to WiFi Effect on Download Speed

Clarifications

I’m not an expert on WiFi,

In reply to WiFi Effect on Download Speed

but to best I can recall off the top of my head, the major difference between
802.11g and 802.11n were in terms of range, and frequency rather than
the speeds…for now. In the future, 802.11n may support up to 100 mbps.
As for why the apparent loss, probably due to signal strength…again, that’s
just my best guess.

WiFi is always slower than a Wired connection

In reply to WiFi Effect on Download Speed

Also if you have any Security Enabled like WEP etc that cuts into the available bandwidth as it’s constantly transmitting the Security Protocols which get included in the Data Transfer so others are unable to access the data stream.

Of course if you are not using any Security and you have anything within half a Kilometer you are an idiot as the likelihood of having your WiFi and Data Hijacked it right over the top.

So the more Secure that the WiFi Security is the slower the Data Transfer is going to be as each Security Measure adds more to what is being Transmitted in each direction and with WiFi it is always 2 way communication where the computer is calling the WiFi Access Point or whatever and requesting whatever and the WiFi Access Point is delivering whatever was requested and doing Data Checks to prevent and Data Corruption. With a weak signal and high security it’s possible that up to 80% of the available bandwidth is being consumed by the Security and Data Checking components of the Wireless link so you only have 20% of the available bandwidth available for Data Transfers.

Yep that’s an Extreme Case but Wireless Signals propagate in different ways and there are way too many ways for them to cause problems things like a Brick Wall, Florescent Light, Microwave Oven or even a Cordless Phone can adversely affect a WiFi Connection.

Col

Reponse To Answer

In reply to WiFi is always slower than a Wired connection

Thanks very much. I do understand that signal strength, security, and other interference like baby monitors can drastically reduce throughput but I’m trying to find out the expected loss in download speed based on a 802.11g connection versus an 802.11n connection. All else being equal. Two clients in the exact same close location to the AP, no security, no other appliances or neighboring networks. One client connected wirelessly at 54Mbps one at 150Mbps. Download speeds are much lower on the G client connected at 54Mbps than the N client connected at 150Mbps, which seems perfectly logical to me. I’m just wondering if there are standard expected percentages of loss.

You are only as fast as the slowest segment

In reply to WiFi Effect on Download Speed

There is more overhead with wireless too so the same transfer speed will deliver more of what you call data over a wired connection than using wireless. The other thing is that wireless is constantly negotiating speed. If it has too many retrys at 54 Mb it will cut in half, then half again and so on until it gets good transmission. After a number of good transmissions it will double the speed and so on back and forth all the time. So your so called 54 Mb connection may really be a 2 Mb connection.

This might help to get your answers

In reply to WiFi Effect on Download Speed

http://www.epitiro.com/assets/files/WiFi_In_The_Home_Epitiro_V2.pdf

Reponse To Answer

In reply to This might help to get your answers

Thanks Samarpatel! This is a great document that does conclude that you can expect about a 30% loss on what appears to be a 802.11g connection versus a wired ADSL connection. I wish Epitiro had an updated case study for 802.11n versus cable, but this is a big piece of what I was looking for. Thanks again.

I have seen some cross talk problems as well on wireless

In reply to WiFi Effect on Download Speed

Where something reflective is reflecting the signal backwards, and thus the laptop gets the same packets twice, out of order. When this happens, your laptop has to request those packets again. Often just rotating the laptop can fix these issues.

I should have titled this thread, 802.11g D/L speeds vs 802.11n D/L speeds

In reply to WiFi Effect on Download Speed

That’s my real question. I do understand the effects of WiFi in general on download speeds.

But if all those effects are equal for both clients, what can you expect in terms of percentage of download speed lost when using a G client as opposed to an N client. Simply because of the difference in link speed (54 vs 150)

There has to be a study on this. No?

Reponse To Answer

In reply to I should have titled this thread, 802.11g D/L speeds vs 802.11n D/L speeds

That’s the problem they are not equal.

The faster the supposed Data Transfer the more susceptible to interference things are and the slower things can be moved around in real life.

Reponse To Answer

In reply to I should have titled this thread, 802.11g D/L speeds vs 802.11n D/L speeds

@HAL 9000
That’s interesting. Ok, so the N client is more susceptible to interference. So if the expected loss in download speed is about 10% for N and 30% for G without interference, you’re saying the loss may end up being 20% for N and only 35% for G when interference is introduced?

Reponse To Answer

In reply to I should have titled this thread, 802.11g D/L speeds vs 802.11n D/L speeds

Way too many variables but that is one possibility.

You would have to take each instance on a location by location or Case by Case Basis. 😉

Col

Reponse To Answer

In reply to I should have titled this thread, 802.11g D/L speeds vs 802.11n D/L speeds

Come on HAL, I’ve seen your other posts in these forums. You’re a smart guy. 😉 Humor me. If all the variables were exactly the same EXCEPT the link speed due to G or N clients, what would be the expected loss in download speed.
There has to be a starting point before the other variables play into it right?
OR is it just a matter of maximum throughput so that UP TO a certain amount of download speed (say 22Mbps) can be successfully handled by both G or N clients and then the G clients max out?? I’ve considered this, but in my experience and testing this doesn’t hold any water.

I refuse to believe that the answer to this question is just a bunch of vague references to cordless phones and baby monitors.

Do any other members of these forums come to mind as being an expert in WiFi that I could PM?

Thanks for the replies!

[Author]

Replies