Does 802.11B get better range than 802.11G?

[MWink]

I've been using wireless for a while now because they forgot to run ethernet wires to my room and I've experimented with quite a bit of different WI-FI hardware. My oldest stuff is some Microsoft 802.11B components, the only piece still in use is the router. I've tried quite a few 802.11G routers but for some reason I get worse reception with them. At first I thought maybe they weren't as good but now I'm starting to wonder if maybe 802.11B just gets better reception?

The routers I've tried include the Linksys WRT54G, the D-Link DI-524, and now the Linksys WRT54GC (compact) and ALL of them have gotten worse reception than my old MS 802.11B router. I've have placed them all in the same place in our "network closet" and toyed with the antenna's (on the one's that had them) but nothing seems to help.

With the MS B router I get 4-5 bars in my room and the signal doesn't start to have problems until I get to the end of the kitchen (the most extreme opposite end of the house from the router). With the G router's (currently the compact Linksys) I get 2 bars in my room and have an unusable signal in the dining room.

I'm wondering if I should just give up and stick with the trusty old MS B router as I guess it is fast enough for most things. I would just like to have a G router since most of my cards are G now. I mainly want to know WHY none of the G routers get reception as good as the B router? Is this normal or do I have a super good B router ? BTW the cards I'm using are a Linksys WUSB54G V4 for my desktop in my room and the Apple Extreme G (or whatever) in my iBook.

 

[JackMDS]

No.

:sun:

 

[ScottMac]

They're on the same frequencies, so it's not an RF thing.

What it is is that there's a higher data density on the signal, and any glitches, noise, or other interference knocks out more bits at higher "speed" (802.11g) than the lower "speed" (802.11g).

802.11g uses a different encoding scheme which tends to tolerate noise better, but everything has its limits.

Good Luck

Scott

 

[MWink]

Any other ideas? It still doesn't make much sense.

 

[mjia]

Have you tried actual bandwidth comparisons? The wireless signals bars are not really a very good indication of performance.

I've run tests on my 802.11g network and in locations with only 35% strength, I get average throughput 13 MB/s, which is obviously even beyond the maximum theoretical of 802.11b.

 

[skyking]

Either one has a finite range, but my experience with G equipment is the same as the poster above.
Even when the signal is getting nearly unusable, I see higher throughput on a G setup.

 

[JackMDS]

Given every thing, equal 802.11b would not providing more Bandwidth then 802.11g

Given your case, and assuming that every thing but the Router is equal.

Read the specs. might be that the ?compact?s? output is smaller than the 802.11b. E.g., the 802.11b is 40mw and the ?compact? is only 30mW

Might be that the 802.11b has a better Antenna. E.g. the ?Compact? is 2dbi the other 4 or 5 dbi.

Might be that your specific ?Compact? unit is ?Lame?.

In any case, as mentioned above to test the actual Bandwidth the Signal Bars in the Entry Level Wireless system mean very little, they can serve as a general indication reference within a unit but not among different units.

:sun:

 

[MWink]

Well regardless of what the reception indicator says the MS router seems to work farther away than the Linksys. BTW how can I test my throughput? I can run an internet throughput test but that would not even come close to saturating an 802.11B network, much less a G network.

 

[mjia]

You can use a program like TTCP.

 

[Kelemvor]

Could also try varying the channel if you haven't already tried that. Try something low, then try something high. See what works best.

And quite frankly, as long as you have bars, you're on. Doesn't really matter how many you have since the connection is faster. We have people at work all the time complaiing that they have yellow or red bard isntead of green ones even though they still have a perfect signal and can do everything they need to do. The bars are just an indicator of the strength of signal, but it still works with a low one most of the time.

 

[Crusty]

Originally posted by: mjia
Have you tried actual bandwidth comparisons? The wireless signals bars are not really a very good indication of performance.

I've run tests on my 802.11g network and in locations with only 35% strength, I get average throughput 13 MB/s, which is obviously even beyond the maximum theoretical of 802.11b.

That's funny because I was unaware that 802.11g was faster then 100mbps networks. I assume you meant 13mega-BITS/s not mega-BYTES.

 

[mjia]

Originally posted by: Crusty

Originally posted by: mjia
Have you tried actual bandwidth comparisons? The wireless signals bars are not really a very good indication of performance.

I've run tests on my 802.11g network and in locations with only 35% strength, I get average throughput 13 MB/s, which is obviously even beyond the maximum theoretical of 802.11b.

That's funny because I was unaware that 802.11g was faster then 100mbps networks. I assume you meant 13mega-BITS/s not mega-BYTES.

Ya...another MB/Mb typo. Thanks for clarifying.

 

[cmetz]

802.11g's lower data rates are to step down to 802.11b when signal is weak, as far as I can tell.

In practice, I think a lot of older 802.11b gear is simply better made than today's SOHO 802.11g gear. My Lucent Orinoco equipment gets a very usable link where my Netgear/Linksys 802.11g equipment can't see one. The Orinoco/WaveLAN cards had much better RF. But then again, you paid for it.

 

[JackMDS]

A 40mW 802.11b would fair better than a 30mW 802.11g.

:sun:

 

[Gillbot]

the WRT54GC is crap. My buddy had one and he could barely get a useable signal out of the room it was in.

 

[nweaver]

Originally posted by: Gillbot
the WRT54GC is crap. My buddy had one and he could barely get a useable signal out of the room it was in.

That is a broad statement for a single unit that has gotten generally good feedback from users. I set up a plain 54G and had reception all over this customers house.

 

[Gillbot]

Originally posted by: nweaver

Originally posted by: Gillbot
the WRT54GC is crap. My buddy had one and he could barely get a useable signal out of the room it was in.

That is a broad statement for a single unit that has gotten generally good feedback from users. I set up a plain 54G and had reception all over this customers house.

The WRT54G is a good unit, I have two of them, I have yet to see a single good review on the WRT54GC. The WRT54GC is the compact unit which doesn't have an external antenna and suffers from random and often dropped connections and needs to be rebooted often. This judgement is based on my personal use as well as quite a few friends experience with that particular unit. Online reviews are pretty harsh for the WRT54GC as well.

 

[Emulex]

Always install the client (wireless cards) as JAPAN
Always use hacked wrt54g firmware
always use G MODE ONLY. NO B
Always use Channel 13.
Up the power output and buy those cheap hawking antenna's.

That formula above is not 100% legal in the usa. woohoo you wont cause any planes to like crash. but i guarantee if you move to channel 13 G-only, you'll see huge gains.

why? All the neighbors poop isn't on 13

p.s. since WEP 128 is weak, use WEP-64 if you dont really care, you'll get more throughput with some encryption. It is just as easy to crack wep-128 as Wep-64 with today's stuff. Otherwise stick to WPA/WPA-2.

no encryption is for the super lazy who really want to share devices.

use MAC address filtering to keep the lamers out. The elite will steal your mac addresses and reflash your router anyways

 

[nweaver]

way to pass along the FUD....lets start at the to

Originally posted by: Emulex
Always install the client (wireless cards) as JAPAN

and hope you don't need A band, right? those are totally off from ours and europes.

Always use hacked wrt54g firmware
always use G MODE ONLY. NO B

so that legacy stuff is broken? Also, not sure with the SOHO gear, but does this cause it to not step down past 11 mb/s when noise/range increases? 5 Mb/s is better then none

Always use Channel 13.

you do realize that 802.11B/G channels overlap by 4, right? that means that if you are on 13, and everyone is on 11, then you still see lots of their packets, but now instead of load on the channel, it's now NOISE on the channel, which can be worse. If you could get to a channel 15 then maybe...

Up the power output and buy those cheap hawking antenna's.

and run the risk of burning out your radios. Also, I help manage an 802.11B internet company, and turning the radio power DOWN helped alot with reliability this winter. Instead of 50% packet loss, we jumped down to 1-5% loss

p.s. since WEP 128 is weak, use WEP-64 if you dont really care, you'll get more throughput with some encryption. It is just as easy to crack wep-128 as Wep-64 with today's stuff. Otherwise stick to WPA/WPA-2.

If your gear actually slows down on wep64 versus wep 128, buy new gear. Most modern stuff doesn't have a noticable slowdown on WPA (TKIP) and barely even measuable with WPA2(AES). Some is not noticable


but thanks for the FUD!

 

[Emulex]

sorry if you consider it fud. This is real world data from real world use. If you have experimented with the WRT54GS in a household situation, you would find the above model for G wireless (who even brought up A? the conversation was G or B) will increase your throughput and signal reliability, and reduce your retrain.

I'm actually moving my core to Corinex AV200 series (for sale now). The Coax will sustain 50-85mbps for dvd streaming over existing coax without interfering with (satellite,cable,cable modem). Or their AV200 series power line which produces less bandwidth.

This will allow me to bridge my network which is in the basement and bring up the other linksys routers and mesh them for strong reliable wireless signal in the house and immediate outside without sacrificing the ability to stream iso/img off smb shares to my xbox for my home movie streaming.

Man you can call it fud, you can throw all the technical jargon around, but if you have not tried it in a real world home setting, then don't dis it. It works. I live in a metro area with at least 8 open wifi's lol, 4 cell phones, 6 (modular) 2.4ghz phones in the house. plus whatever else that throws off the signals. I went from nada very low 10-100%(packet loss) to 10-25% packet loss with the latter being when a proximity mobile phone or cordless was activated (i picked up the cordless sitting next to the wrt54gs).

YMMV.


If i had the $$ to buy 5.8ghz A band i probably would. But since my laptops would require new cards pcmcia, no minipci. (2 laptops), my pda (doesnt support A), and my router doesn't support A i'm not going to move to another band.

I am aware of the channels. I have tried, 1,2,3,4,5,6,7,8,9,10,11,12,13 (not 14-15 not sure if my gear supports that) and found for my situation 13 was the best. I did this systematically because i wanted to see the affects of mobile phone proximity, cordless phone proximity, microwave and other devices that cause the signal to retrain.

And yeah since there are alot of folks using legacy B in the neighborhood on their fancy G routers, you get alot of crap blasted.

What i see is 18-36mbps rates, with less retrains than ever..

costs nothing to try it. When i say up the power, i am not saying set your linksys to power 242 with the stock radio antenna's. I think we all know the consequences of that. Everything in moderation.

 

[nweaver]

great for your real world settings. All I'm saying is that the tech goo behind is the adjacent 2 channels share frequency space, so 1 shares with 2 and 3, 2 with 1, 3, and 4, etc. 13 shares with 11 and 12 (not really a 14 and 15)

When we are testing, and need to generate 2.4 Ghtz noise for noise histogram reports (part of the Cisco wireless standards) know what we use? Another client, on another AP, 2 channels away. Works like a charm. One channel away is seen as some noise, but mostly load on the channel. So your noise goes up (there are lots of ch11 users out there) and your channel load goes down.


just as an fyi, I DON"T have a lot of home experience. I live in the middle of nowhere, and there are maybe 10 access points between me and the highway. I DO however, sit in a faraday cage testing wirless equipment for compliance to the Cisco Compatible Extensions program, so I'm not exaclty googling this stuff

 

[Emulex]

yeh i'm sure we both hacked our 300 baud modems to 450 baud back in 78 but that's not really helping the dude.

what are your suggestions as a total plan. I will be glad to put your non-googled suggestions to implementation in my network and give you a subjective (but not anlaytical) comparison.

ok? Tell me how to setup my gear, and i'll try it. You probably are more intelligent than I am. I'll bite. go for it.

wrt54gs v.2 sveasoft, basement only location, 2 hawking non directionals on linksys powered at about 100 out of 255, G-only, client pc 2 stories up with hawking antenna. 1 pda (not used hardly), and 2 laptops G-only.

tell me what is ideal given constraints on the environment

 

[nweaver]

1. Replace gear with enterprise level (I've not really worked on the home stuff). CB21 adapters (or atheros based miniPCI) for clients, cisco1200 series AP's. Get 2 of them
2. Setup AP's to be managed by the WLSE server
3. ...
4. Profit?



I'm not saying that will work for a home (too much $$) but using channel 13 does NOT eliminate noise caused by devices on adjacent channels, or 2.4 Ghz phones. Encryption DOES NOT slow down your wireless connection a significant amount. (trust me, I've run well over 70 different cards through testing, and that's one test case, throughput with different security levels).