802.11n Field Test

[Peel]

OK, so I downloaded the 802.11n updater for my MacBook this afternoon. Before installing, I did some test file transfers, so I could compare the speed differential between N and G.

Here's the setup:

equipment:
Linksys WRT300N router
iMac G5 Gigabit Ethernet connection to Router
MacBook C2D connecting via airport card

Test File: 59.2MB QuickTime h.264 video

Test distances:
test 1: 1 meter distance, no obstructions between MacBook and Router
test 2: 20 meters distance, wood and plaster walls between MacBook and Router

I tested an upload from MacBook to iMac, and a download of the same file from iMac to MacBook. Each test was done 4 times, and the numbers shown below are averages. I ran the test in a mixed mode on router before the upgrade, in N-only mode on router, and mixed-mode after the upgrade, for a total of 48 file transfers.

Transfer times are in seconds

Test 1: 1 meter, no obstructions

MacBook Router Upload Download
.....G......Mixed....40.........35
.....N.........N.......31.........12
.....N......Mixed....31.........20

Test 2: 20 meter, with obstructions

MacBook Router Upload Download
.....G......Mixed....65.........62
.....N.........N.......31.........12
.....N......Mixed....32.........19

Most interesting to me was the disparity between up and downloads. I wouldn't have thought there to be a difference here, and I'm not sure where this is coming from: the MacBook, Router, or iMac.

Also notice the time difference for the long distance test in G versus the same tests in N. I was sitting on my front sidewalk for this test, and in G mode (pre-upgrade) I had 1-2 bars showing on the Airport icon. In N mode I had all 4 bars showing all the time. This alone is worth the price of admission to me

 

[micsaund]

Wow - 48 xfers - now that's a test!

Thanks for taking the time to do that and share the results. It definitely looks like N makes a real, noticeable difference.

Mike

 

[Peel]

Wow - 48 xfers - now that's a test!

Well, I'm always skeptical of someone who says "I got such-and-such time" when they only tried it once. There are too many variables that can sway the results, and you just don't know if you only do one of each test, therefore I did four for each scenario.

 

[Abstract]

Thanks for doing all that work.

 

[NickFury]

Whats really cool to see too is how 'N' is handling obstructions much better. The times didn't seem to change much with or without obstructions when using N.

I've been on the edge about upgrading from my trusty G, but I don't get very good signal around my house, and it seems like N would really make a difference.

 

[adiosk8]

that is a small file to use to test, you would need something that is atleast 500 MB to do something more accurate

 

[djdawson]

that is a small file to use to test, you would need something that is atleast 500 MB to do something more accurate

As long as the transfer protocol has time to reach a relatively steady state it's not necessary to use huge files. If his tests were running for several seconds that's probably good enough, especially if that's the kind of file he regularly deals with. There's no single "best" file size for network benchmarking, just as there's no single "best" app or set of apps for doing system benchmarks - use what's comparable to what's important to you.

 

[BlackMax]

Thanks for the field test. Looks like "N" is a good improvement over "G". I wonder how long it will take my favorite hotspots around town to upgrade their APs to N?

 

[smacman]

Question for you Peel. How do you like that router? I am seriously considering getting it instead of the Airport Extreme but I need to know that it is stable. I had a Belkin N router which caused me nothing but headaches with iChat and constant dropped connections so I am a little gunshy now.

Thanks for doing all that test work. You have me sold on 802.11n.

Cheers

 

[copanewbe]

Thanks for the field test. Looks like "N" is a good improvement over "G". I wonder how long it will take my favorite hotspots around town to upgrade their APs to N?

Are you going to be trading a lot of files over the networks in public places? The only reason I say this is because people seem to think that N is going to make browsing the internet a lot faster, but it's not going to do that much at all.

I am impressed with its ability to go through ubstructions. Really good work and nice test overall, but I'll stick to Gigabit ethernet for networking .

 

[bbydon]

what about battery time on the laptop.
Does the N drain more power?

 

[erikpmort]

what router do you need? i have a D-link router but i think it just does G... do you have to buy a new one to do n ?

 

[e12a]

what router do you need? i have a D-link router but i think it just does G... do you have to buy a new one to do n ?

Yes.

 

[mkrishnan]

Wow, nice work! Thank you for sharing your results!