Data Transfer Rate for 54Mbps 3com link

I've just measure it & find I'm getting an effective transfer rate of only 1.67MB per second. Seems a bit on the low side? What sort of speeds are deemed acceptable by the cognoscenti around these parts?

Reply to
Nho Whei
Loading thread data ...

maybe it's just today?

-------------------------------------------------------------------------------------------------------------------------

POKER ACTION AT

formatting link

Reply to
AZApoker

1.67 MB/s (~13 Mb/s) on a 54 Mb/s link is pretty decent if the data source and the data sink are both on the same 802.11G segment.

If you meant 1.67 Mb/s, then something is wrong.

Reply to
Bob Willard

On the face of it Bob, ~13Mb/s doesn't stack up too well against

54Mb/s though does it, or am I missing something too?
Reply to
Nho Whei

Nho Whei hath wroth:

You're missing something.

Under ideal conditions, a 54Mbit/sec wireless connection will give a maximum of about 25Mbits/sec thruput. See the FAQ:

formatting link
a table of speeds. If you have the 802.11b compatibility mode enabled, your maximum speed will be about 14Mbits/sec, which is apparently what you're getting.

If you're going to run benchmarks, I suggest you supply some clue as to what hardware you're using and how you're testing. To get

25Mbits/sec thruput, you need to be fairly close as the wireless connection speed will automagically decrease with increased range. The test should be performed between a wired computer acting as a server, the access point or router, and the wireless connection. If you try it wireless to wireless, you will again get half the speed or perhaps 12Mbits/sec max. I suggest using Iperf 1.7:
formatting link
Reply to
Jeff Liebermann

Do you mean "802.11b compatibility mode enabled", or in use? I see iperf speed of 26.4 Mbits/sec between 54g and a wired desktop, and my WAP is capable of communicating with 802.11b, although none are connected right now. iperf duplex shows 13.1 + 14.6 Mbits/sec

I can write to a Windows share at 2461 KBytes/Second.

Maybe our 14Mbits/sec poster has some 802.11b devices connected to his WAP.

Reply to
dold

snipped-for-privacy@XReXXDataX.usenet.us.com hath wroth:

Connected, but no traffic. See: |

formatting link
Since the key feature of 802.11g is backward compatibility with 802.11b, throughput tests should be done with an 802.11b client device connected to the access point but otherwise idle. This setup ensures that the 802.11g network is operating in an 802.11b compatible mode. When I tried it with a Broadcom based WRT54G, it didn't quite work that way.

If the 802.11b compatibility mode is enabled, the access point spends a small amount of time listening for 802.11b traffic. It apparently (from my testing) can can be *ANY* 802.11b traffic and does not need to be the same SSID. When it hears even one 802.11b packet, it enlarges the sampling time so that it can properly decode the 802.11b packets. I think (not sure) that the time slicing is about 2/3

802.11g and 1/3 802.11b.

I ran the usual IPerf tests on my WRT54G v1.1 and got something near the expected 25Mbits/sec with 802.11b compatibility turned off. When I turned it on, the speed slowed down to about 22Mbit/sec. I then fired up my Orinoco Silver card on another laptop and connected to the WRT54G. Thruput dropped to about 13Mbits/sec. Note that there was no real 802.11b traffic, just a connection. When I dropped the connection, the speed improved dramatically after about 15-30 seconds.

I must admit that I did a really sloppy job of traffic measurement and that the numbers are from my foggy memory. I'll do it again when my WRT54G returns from a site survey.

Reply to
Jeff Liebermann

I see 26M with my Netgear WRG614 set for 802.11b compatibility, and no

802.11b clients connected.

I haven't tested with the compatibility mode turned off.

I plugged in an 802.11b USB dongle, whcih connected. No traffic on

802.11b. My rate dropped to 16MBits/Sec.

So, it's okay to have the compatibility mode turned on for backwards compatibility, but actually having the 802.11b clients connected will slow down the 54g devices, with performance returning when the b device goes away.

Reply to
dold

On Fri, 25 Aug 2006 19:08:51 +0000 (UTC), snipped-for-privacy@XReXXDataX.usenet.us.com wrote in :

How about with it plugged in but not connected?

You haven't presented enough evidence for that conclusion.

Reply to
John Navas

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.