Wireless vs Wired connection

[integramodder]

This is the wireless connection:
Intel(R) PRO/Wireless 3945ABG Network Connection
~30 feet from router
http://www.speedtest.net/result/569482186.png


This is the wired connection:
Marvell Yukon 88E8036 PCI-E Fast Ethernet Controller
~30 feet of cable
http://www.speedtest.net/result/569484278.png

Network:
D-Link Gigabit Switch
Linksys WRT54G - DD-WRT v24-sp2 (05/21/09) std - build 12188M NEWD Eko
Wireless secured with 64 Bit WEP


Why is the wireless download speed half as much are the hardwire?

 

[xSauronx]

wireless, due to signal loss and physical/radio interference just cant sustain the same kinds of speeds that you can get on a wired network...and you never said if it was 30 feet away from the router, but you can see the router, or if it was 30 feet away from the router through a wall or what not. it can go faster than the 10Mbps youre showing, but due to the previously mentioned factors, the throughput can take a beating pretty easily.

mimo and wireless n can help with that some, but wireless is still not going to be as good as wired.

 

[frostedflakes]

Keep in mind that due to overhead, you won't get transfer speeds anywhere close to 54Mbps. Typical throughput for 802.11g is about 20Mbps, though, so you should be getting better speeds than that. How's the signal strength? It might be dropping you down to lower speeds, which will limit throughput.

edit: Also make sure you don't have any other wireless devices operating in the 2.4GHz band that could be interfering (cordless phones, mice/keyboards, etc.).

 

[Emulex]

plus latency and retrain latency and authentication latency when a connection fails completely.

wireless is worst than powerline networking; think of it as a wireless modem; do you remember the 56k days?

it's great for light use however. commercial gear is a bit more stable and well planned.

 

[Engineer]

Tomato is faster!



Edit: Damnit, that's only LAN to WAN tests for Tomato....oops!

(Please ignore as I can't read)!

I just did a PC to PC transfer using wireless and it topped out at around 2MB/sec. Not great. It does max out my broadband connection of 10Mbps (actually, its the fastest of the PC's in the house by 100Kbps...not sure why! )

 

[Lemon law]

The other thing is to note is the fact that hard wired network does not need to security programs that a wireless network does.
Because without the wireless security added, anyone close by can join your network.

 

[integramodder]

Originally posted by: xSauronx
wireless, due to signal loss and physical/radio interference just cant sustain the same kinds of speeds that you can get on a wired network...and you never said if it was 30 feet away from the router, but you can see the router, or if it was 30 feet away from the router through a wall or what not. it can go faster than the 10Mbps youre showing, but due to the previously mentioned factors, the throughput can take a beating pretty easily.

mimo and wireless n can help with that some, but wireless is still not going to be as good as wired.

Its probably more like 25 feet line of sight.

Originally posted by: frostedflakes
Keep in mind that due to overhead, you won't get transfer speeds anywhere close to 54Mbps. Typical throughput for 802.11g is about 20Mbps, though, so you should be getting better speeds than that. How's the signal strength? It might be dropping you down to lower speeds, which will limit throughput.

edit: Also make sure you don't have any other wireless devices operating in the 2.4GHz band that could be interfering (cordless phones, mice/keyboards, etc.).

http://i33.tinypic.com/5k4pb5.jpg
Its showing 36 Mbps. I don't know why its not 54, either the router or Intel card.

Originally posted by: Emulex
plus latency and retrain latency and authentication latency when a connection fails completely.

wireless is worst than powerline networking; think of it as a wireless modem; do you remember the 56k days?

it's great for light use however. commercial gear is a bit more stable and well planned.

I'm not expecting to get hardwired like connection over wireless, but a reduction to half the download speed seems like something is wrong.

Originally posted by: Engineer
Tomato is faster!



Edit: Damnit, that's only LAN to WAN tests for Tomato....oops!

(Please ignore as I can't read)!

I just did a PC to PC transfer using wireless and it topped out at around 2MB/sec. Not great. It does max out my broadband connection of 10Mbps (actually, its the fastest of the PC's in the house by 100Kbps...not sure why! )

Those tests were also all done on wired connections. I love DD-WRT over the standard Linksys firmware, but I haven't done much research on Tomato, need to do that...

Originally posted by: Lemon law
The other thing is to note is the fact that hard wired network does not need to security programs that a wireless network does.
Because without the wireless security added, anyone close by can join your network.

I know, as I said above, I don't expect it to match wired, but the difference is too great to ignore. I suppose I could switch to MAC filtering and turn off WEP to see if there is a notable speed improvement.

 

[spidey07]

That's totally normal for wireless (802.11g in particular). It's slow. Nothing's really wrong, what you're seeing is normal. Read the sticky about channels, noise, antenna positioning and interference.

It's doubtful that your router is doing WEP in software and not hardware. It's possible with some ancient models, but highly unlikely.

 

[Engineer]

Originally posted by: integramodder


Those tests were also all done on wired connections. I love DD-WRT over the standard Linksys firmware, but I haven't done much research on Tomato, need to do that...

Yes, I noticed that AFTER I had posted and even edited as much to reflect that I cannot read, lol. Sorry about that!

 

[integramodder]

Originally posted by: spidey07
That's totally normal for wireless (802.11g in particular). It's slow. Nothing's really wrong, what you're seeing is normal. Read the sticky about channels, noise, antenna positioning and interference.

It's doubtful that your router is doing WEP in software and not hardware. It's possible with some ancient models, but highly unlikely.

Have already read the sticky and I did most of that when setting up my network...

 

[spidey07]

Originally posted by: integramodder

Originally posted by: spidey07
That's totally normal for wireless (802.11g in particular). It's slow. Nothing's really wrong, what you're seeing is normal. Read the sticky about channels, noise, antenna positioning and interference.

It's doubtful that your router is doing WEP in software and not hardware. It's possible with some ancient models, but highly unlikely.

Have already read the sticky and I did most of that when setting up my network...

If you've done all that then I'm inclined to think you just have normal wireless throughput for 802.11b/g. It depends on your environment.

 

[Qrilock]

In addition to the wireless interference, additional overhead, and the fun of the entire area being a single collision domain, wireless is a half duplex technology. It would be more realistic to compare your download speed against a wired network running half duplex through a loaded hub with poor cabling.

 

[spidey07]

Originally posted by: Qrilock
In addition to the wireless interference, additional overhead, and the fun of the entire area being a single collision domain, wireless is a half duplex technology. It would be more realistic to compare your download speed against a wired network running half duplex through a loaded hub with poor cabling.

And right next to a huge EM field from a nuclear generator. That might be a fair comparison.

 

[imported_hopeless]

Just another possibility.

I can't seem to find the site to link, but awhile back (over a year) I found a site that did wireless tests using: Wide open, WEP, WPA/TKIP, and WPA/AES. The results showed a loss when using WEP or WPA/TKIP, but with WPA/AES it was right back up to where it was with no security at all.

I'll see if can find it and post a link.

 

[Engineer]

Originally posted by: hopeless
Just another possibility.

I can't seem to find the site to link, but awhile back (over a year) I found a site that did wireless tests using: Wide open, WEP, WPA/TKIP, and WPA/AES. The results showed a loss when using WEP or WPA/TKIP, but with WPA/AES it was right back up to where it was with no security at all.

I'll see if can find it and post a link.

IIRC, that's because the WEP hardware has been removed and it's being done in software now, causing the router CPU to eat up resources. WPA/WPA2 in hardware replaced WEP. Not 100% that this is the case, but I have heard it (probably from this very forum! )