average data rates

After searching over dozen of sites, I've found followind average data rates
for 802.11x:
802.11a (54 Mbps) - 27,3 Mbps
802.11b (11 Mbps) - 5,6 Mbps
802.11g (54 Mbps) with RTS/CTS protection - 8,8 Mbps
802.11g (54 Mbps) without any protection - 27,3 Mbps

Could this be true?

And does anyone knows what is max. distance between 2 stations or APs for
that standards, without using any amplifiers or external antennas??

tnx in advance

In article <bqojp4$lpg$(E-Mail Removed)>,
John_2001 <(E-Mail Removed)> wrote:
:After searching over dozen of sites, I've found followind average data rates
:for 802.11x:
:802.11a (54 Mbps) - 27,3 Mbps
:802.11b (11 Mbps) - 5,6 Mbps
:802.11g (54 Mbps) with RTS/CTS protection - 8,8 Mbps
:802.11g (54 Mbps) without any protection - 27,3 Mbps

:Could this be true?

I was recently looking over the reviews at tomshardware and
smallnetbuilder . The highest throughput I saw for any 802.11g device
was about 22 1/2 Mbps, and most where 18-20 Mbps.

:And does anyone knows what is max. distance between 2 stations or APs for
:that standards, without using any amplifiers or external antennas??

There is no inherent maximum distance to the -standards-. There are,
though legal limitations in some jurisdictions that impose practical
limits.
--
Cannot open .signature: Permission denied

"John_2001" <(E-Mail Removed)> wrote in message
news:bqojp4$lpg$(E-Mail Removed)...
> After searching over dozen of sites, I've found followind average data
rates
> for 802.11x:
> 802.11a (54 Mbps) - 27,3 Mbps
> 802.11b (11 Mbps) - 5,6 Mbps
> 802.11g (54 Mbps) with RTS/CTS protection - 8,8 Mbps
> 802.11g (54 Mbps) without any protection - 27,3 Mbps
>
> Could this be true?
>

If anything, they are on the high side. These numbers mostly just reflect
the fact that because 802.11 is half-duplex (can't transmit and receive at
the same time) the raw data rate is cut in half.

If you are concerned about actual data throughput, you have to knock off
even more for packet overhead, etc. So, for example, the effective
throughput of 802.11b is only about 4 Mbps and change.


> And does anyone knows what is max. distance between 2 stations or APs for
> that standards, without using any amplifiers or external antennas??
>

There are far too many variables for a simple anwer to that question. You
have to make assumptions about Tx output, Tx antenna gain, Rx antenna gain,
Rx sensitivity, and the desired modulation/speeed. Then factor in the
environment: open-air, indoor construction, height above ground, etc.

Some links have been made at tens of miles; some links fail at ten yards.

On 2003-12-05, John_2001 <(E-Mail Removed)> wrote:
> After searching over dozen of sites, I've found followind average data rates
> for 802.11x:
> 802.11a (54 Mbps) - 27,3 Mbps
> 802.11b (11 Mbps) - 5,6 Mbps
> 802.11g (54 Mbps) with RTS/CTS protection - 8,8 Mbps
> 802.11g (54 Mbps) without any protection - 27,3 Mbps
>
> Could this be true?
>
> And does anyone knows what is max. distance between 2 stations or APs for
> that standards, without using any amplifiers or external antennas??
>
> tnx in advance
>
>
My experience tells me this :
802.11a (54 Mbps) - 20 Mbps
802.11b (11 Mbps) - 4,5 Mbps

Greets

--
<< IRC : IrcNet : #bsd4pl, #llug >< WWW : http://darkzone.ma.cx >>
<< Linux Registered User : #212220 >< GG : 3383949 >>
<< "Forever trusting who we are and nothing else matters" >>

"Ross Evans" <(E-Mail Removed)> wrote in message
news:10Rzb.87139$(E-Mail Removed).. .
>
> "John_2001" <(E-Mail Removed)> wrote in message
> news:bqojp4$lpg$(E-Mail Removed)...
> > After searching over dozen of sites, I've found followind average data
> rates
> > for 802.11x:
> > 802.11a (54 Mbps) - 27,3 Mbps
> > 802.11b (11 Mbps) - 5,6 Mbps
> > 802.11g (54 Mbps) with RTS/CTS protection - 8,8 Mbps
> > 802.11g (54 Mbps) without any protection - 27,3 Mbps
> >
> > Could this be true?
> >
>
> If anything, they are on the high side. These numbers mostly just reflect
> the fact that because 802.11 is half-duplex (can't transmit and receive at
> the same time) the raw data rate is cut in half.

802.11 protocols are half-duplex, but that has nothing to do with these data
rates. Max throughput is measured by a one-way data transfer from a wireless
client to or from a server directly connected to an Ethernet port on the
router. See http://www.atheros.com/pt/atheros_be...whitepaper.pdf. The
gap between max theoretical bandwidth and actual bandwidth is due entirely
to 802.11 protocol overhead and TCP/IP protocol overhead. According to

http://www.atheros.com/pt/atheros_range_whitepaper.pdf

the maximum throughput for 802.11g using UDP/IP traffic with 1500 byte
payloads and WEP enabled is 30.5 mbps, and for 802.11 b it's 7.1 mbps. These
are more than half the theoretical max throughputs. The numbers are from an
analytical probability model, and real measured numbers are always less, but
they make the point.

If you transmit data between wifi clients in an infrastructure network,
where all data is bridged through the AP, then the throughput is cut in half
as you say. Each frame must be completely received and then retransmitted.
So, a real throughput of 22 mpbs would become 11 mbps. This would be true
even if the AP had two antennas. For discussion's sake, say that you have an
AP that bridges between an 802.11g and an 802.11a network. It can
simultaneously receive on one network, and transmit on the other, so it is
not necessarily a half-duplex arrangement. Even so, a bridge is not allowed
to retransmit an errored frame. So, it is obliged to completely receive
every frame before retransmit, and discard any frame in error. This
restriction would remain even if 802.11 were redesigned to be full duplex -
the AP would not be allowed to retransmit a frame in error. In an ad-hoc
network, stations communicate directly. One-way transfers between stations
can exceed half the theoretical max.

>
> If you are concerned about actual data throughput, you have to knock off
> even more for packet overhead, etc. So, for example, the effective
> throughput of 802.11b is only about 4 Mbps and change.
>
>
> > And does anyone knows what is max. distance between 2 stations or APs
for
> > that standards, without using any amplifiers or external antennas??
> >
>
> There are far too many variables for a simple anwer to that question. You
> have to make assumptions about Tx output, Tx antenna gain, Rx antenna
gain,
> Rx sensitivity, and the desired modulation/speeed. Then factor in the
> environment: open-air, indoor construction, height above ground, etc.
>
> Some links have been made at tens of miles; some links fail at ten yards.
>
>

How can one determine the actual data rates for their setup? I have
just sucessfully put in a 1/4 mile wireless link using two WAP54g
access points. I am getting no packet loss and 1.8 ms ping times.
However, I am not sure how to measure the real throughput of the link.
Is their any software that can be used to measure the rate?

Thanks,

Fred


>>
>
>If anything, they are on the high side. These numbers mostly just reflect
>the fact that because 802.11 is half-duplex (can't transmit and receive at
>the same time) the raw data rate is cut in half.
>
>If you are concerned about actual data throughput, you have to knock off
>even more for packet overhead, etc. So, for example, the effective
>throughput of 802.11b is only about 4 Mbps and change.
>
>
>> And does anyone knows what is max. distance between 2 stations or APs for
>> that standards, without using any amplifiers or external antennas??
>>
>
>There are far too many variables for a simple anwer to that question. You
>have to make assumptions about Tx output, Tx antenna gain, Rx antenna gain,
>Rx sensitivity, and the desired modulation/speeed. Then factor in the
>environment: open-air, indoor construction, height above ground, etc.
>
>Some links have been made at tens of miles; some links fail at ten yards.
>

Here's a couple of internet performance test links:

http://www.pcpitstop.com/
http://www.dslreports.com/stest

I prefer the dslreports tests. They offer 3 free test sites, and an optional
list of 212 free sites worldwide that you can use if the others are busy.

These test performance from your host to an internet server - not local
throughput between wifi clients, or throughput to a directly connected
server. The dslreports sites claim to be connected directly to the internet
backbone via DS3 lines. I recently tried about a dozen of these sites, and
got amazingly consistent results, even though I chose servers in several
different states, and tried at different times of day.

Your results may be less than optimal. The best you can say from such tests
is that your internet throughput can be *at least* this good, under the
right conditions. Things that can affect the test:

- how well-connected your local ISP is to the backbone
- how well-tuned your system is
- how busy the servers are
- if you test a cable connection against an ADSL-optimized server, or
vice-versa, you may not get best results - if you have ADSL, you should
probably avoid the RoadRunner sites, or any site that specifically says it's
to test a cable connection (though I didn't see much real difference).

You can check your local IP tuning by looking for sites like
http://www.speedguide.net (run the tcpip analyser).



"Fred J Keller" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> How can one determine the actual data rates for their setup? I have
> just sucessfully put in a 1/4 mile wireless link using two WAP54g
> access points. I am getting no packet loss and 1.8 ms ping times.
> However, I am not sure how to measure the real throughput of the link.
> Is their any software that can be used to measure the rate?
>
> Thanks,
>
> Fred
>
>
> >>
> >
> >If anything, they are on the high side. These numbers mostly just
reflect
> >the fact that because 802.11 is half-duplex (can't transmit and receive
at
> >the same time) the raw data rate is cut in half.
> >
> >If you are concerned about actual data throughput, you have to knock off
> >even more for packet overhead, etc. So, for example, the effective
> >throughput of 802.11b is only about 4 Mbps and change.
> >
> >
> >> And does anyone knows what is max. distance between 2 stations or APs
for
> >> that standards, without using any amplifiers or external antennas??
> >>
> >
> >There are far too many variables for a simple anwer to that question.
You
> >have to make assumptions about Tx output, Tx antenna gain, Rx antenna
gain,
> >Rx sensitivity, and the desired modulation/speeed. Then factor in the
> >environment: open-air, indoor construction, height above ground, etc.
> >
> >Some links have been made at tens of miles; some links fail at ten yards.
> >
>

If you just want to measure just the link performance, I suggest you read

http://www.atheros.com/pt/atheros_be...whitepaper.pdf

It suggests several benchmarking packages, including netperf, which is
widely-used and free.

"Fred J Keller" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> How can one determine the actual data rates for their setup? I have
> just sucessfully put in a 1/4 mile wireless link using two WAP54g
> access points. I am getting no packet loss and 1.8 ms ping times.
> However, I am not sure how to measure the real throughput of the link.
> Is their any software that can be used to measure the rate?
>
> Thanks,
>
> Fred
>
>
> >>
> >
> >If anything, they are on the high side. These numbers mostly just
reflect
> >the fact that because 802.11 is half-duplex (can't transmit and receive
at
> >the same time) the raw data rate is cut in half.
> >
> >If you are concerned about actual data throughput, you have to knock off
> >even more for packet overhead, etc. So, for example, the effective
> >throughput of 802.11b is only about 4 Mbps and change.
> >
> >
> >> And does anyone knows what is max. distance between 2 stations or APs
for
> >> that standards, without using any amplifiers or external antennas??
> >>
> >
> >There are far too many variables for a simple anwer to that question.
You
> >have to make assumptions about Tx output, Tx antenna gain, Rx antenna
gain,
> >Rx sensitivity, and the desired modulation/speeed. Then factor in the
> >environment: open-air, indoor construction, height above ground, etc.
> >
> >Some links have been made at tens of miles; some links fail at ten yards.
> >
>

John_2001 <(E-Mail Removed)> wrote:
> After searching over dozen of sites, I've found followind average data rates
> for 802.11x:
> 802.11a (54 Mbps) - 27,3 Mbps
> 802.11b (11 Mbps) - 5,6 Mbps

In practice.
For 802.11b with only one client with good signal a few meters from the AP
connected to a remote host over 10base2 cable.
(linux 2.6.0-test9, atmel 11b card, linksys WRT54G router)
NFS (disk share) 720K/s.
FTP 710K/s.
1500 byte pings 361K/s.