On Mon, 2006-06-19 at 13:08 -0500, Mikkel L. Ellertson wrote: > James Pifer wrote: > >> General note: scp is great for real-life use over unprotected networks. > >> But because there is the overhead of encryption and decryption, scp can > >> be limited by processor power at one end or the other of a link, rather > >> than raw throughput (especially if the network card itself requires a > >> lot of processor time to do its stuff). This means it's not the best > >> choice for measuring anything other than scp throughput. > > > > Good point. I did not take into consideration the wireless encryption > > being done. Using ttcp I got these results: > > > > 11M = 4.48 Mbit/sec +++ > > 54M = 8.68 Mbit/sec +++ > > > > On a regular 10/100 NIC I get 87.03 Mbit/sec +++ > > > > With or with encryption overhead that is significantly faster than the > > wireless connection. > > > > Any ideas? > > > > Thanks, > > James > > > Dumb question - are there any other users using the same access > point? I would have expected 50M to be faster then what you show, > but I do not expect it to be as fast as the wired connection. > > If I remember correctly, there is more "overhead" in a wireless > connection, and that eats some of the bandwith. Also, other users on > the same access point are sharing the access point bandwith, and > that can cut down the bandwith you can use. You run into the same > problem on a wired connection if you are using a hub, but not if you > are using a switch or router. > > To make matters more interesting, you may also have more processor > overhead with the wireless connection if you are using WEP or WPA. > For a lot of NICs, the encryption/decryption is done in software. > > Mikkel > -- No, no other users. This is my home office network. I have one other machine on the access point, and it's not doing anything. I'm using WEP and MAC filtering. I agree that WEP would certainly add some overhead, that's to be expected, but to only get twice the speed of my 11 meg connection rather than 5x, seems strange. I'm not saying definitely, but couldn't that be more of a driver issue with bcm43xx? Anyone willing to do the same test as I did using ttcp with their wireless connection? Preferably with WEP enabled. I have a non-wireless machine running: ttcp -r -s Then on my machine run ttcp -t [ip address of machine above] -s -f m Then I ran it again after I changed my speed using: /sbin/iwconfig eth1 rate 11M or /sbin/iwconfig eth1 rate 54M Depending on what speed I was running. I'd be interested to see what others get with the same test. Thanks, James
