wireless g speeds - 22 mbps? 54 mbps 108 mpbs???

I have a question regarding wifi speeds. I am buying a wireless usb
adapter for a new tivo and am looking for adapters.

Netgear has a 108 Mbps adater, but I was trying to do some research to
find out if that is a marketing gimmick. I had thought that 54 Mbps
was the maximum thruput, but I was reading on cnet that 802.11 g specs
are:

20Mbps 150 feet range.

How can wifi g adapters give 54 or even 108 Mbps? Is this done through
duplexing?

They say it's compatible with any wireless g device, so I don't think
they're using any kind of proprietary boosting technology.

Any info would be helpful and appreciated.

Thanks,

Seth

On 9/3/05 11:15 AM, in article
(E-Mail Removed) om, "Seth"
<(E-Mail Removed)> wrote:

> I have a question regarding wifi speeds. I am buying a wireless usb
> adapter for a new tivo and am looking for adapters.
>
> Netgear has a 108 Mbps adater, but I was trying to do some research to
> find out if that is a marketing gimmick. I had thought that 54 Mbps
> was the maximum thruput, but I was reading on cnet that 802.11 g specs
> are:
>
> 20Mbps 150 feet range.
>
> How can wifi g adapters give 54 or even 108 Mbps? Is this done through
> duplexing?
>
> They say it's compatible with any wireless g device, so I don't think
> they're using any kind of proprietary boosting technology.
>


I think you have a two questions here:

What speed can I get with a 108 adapter on TiVo?

Answer: about 22 Mbps, because the TiVo doesn't support Super-G mode on that
USB adapter. You need the latest PC driver for that.

Why do network companies use bogus marketing terms?

Answer: 108 and 54 are like the perfect date -- it's not even theoretical
possible, because there are too many variables. But the terms do have an
underlying meaning. It's usually about a quarter to a half for 54, 108 and
125 listed speeds.

> Answer: 108 and 54 are like the perfect date -- it's not even theoretical
> possible, because there are too many variables. But the terms do have an
> underlying meaning. It's usually about a quarter to a half for 54, 108 and
> 125 listed speeds.

54Mbps is the signalling rate, not the data throughput rate. It's not
the same thing.

David.

First, thanks to both of you for your input.

So now I have two questions. One practical question and one question
out of curiosity.

First question is, for Tivo, which adapter do you recommend? Netgear
(or other brand's) 54 Mbps G adapter?

Second question is, what is the difference beteen signalling rate and
thruput rate? Is it that signalling rate is the rate at which data is
being sent out of the usb adapter and thruput rate is the "actual"
effective rate being received (which would be affected by interference,
collisions, dropped packets, etc)?

Thanks!

Seth

On Sat, 03 Sep 2005 09:15:22 -0700, Seth wrote:

> I have a question regarding wifi speeds. I am buying a wireless usb
> adapter for a new tivo and am looking for adapters.
>
> Netgear has a 108 Mbps adater, but I was trying to do some research to
> find out if that is a marketing gimmick. I had thought that 54 Mbps
> was the maximum thruput, but I was reading on cnet that 802.11 g specs
> are:
>
> 20Mbps 150 feet range.
>
> How can wifi g adapters give 54 or even 108 Mbps? Is this done through
> duplexing?
>
> They say it's compatible with any wireless g device, so I don't think
> they're using any kind of proprietary boosting technology.
>
> Any info would be helpful and appreciated.
>
> Thanks,
>
> Seth

All I have to say about this subject is that any wireless G stuff that
goes beyond 54 Mbps is using technology that isn't a standard - in other
words: proprietary. I'd stay away from that stuff even though the
increased throughput is tempting.

As for the 20Mbps rate at 150 feet - the bit rate declines as you get
farther away from the AP.

> Second question is, what is the difference beteen signalling rate and
> thruput rate? Is it that signalling rate is the rate at which data is
> being sent out of the usb adapter and thruput rate is the "actual"
> effective rate being received (which would be affected by interference,
> collisions, dropped packets, etc)?

No, signalling rate is best described as the bit rate while actually
sending bits. Pretty much like the speed at which you talk could be
described as words per minute, however in conversation, you don't talk
continuously without breath, punctuation or waiting to see if anyone
else wants to say anything. If you could talk continuously then the
signalling rate would be the same as the actual throughput but the real
number of words you get out a minute, although spoken at a particular
rate, is not the same, that's your throughput.

Same as on a network, the hardware device signals at a certain rate, on
ethernet it could be 100Mbps, devices receive at the same rate, you
can't have devices sending at any old speed otherwise the timing just
doesn't work. However there are gaps in network traffic so when you
measure actual bytes throughput of a file transfer, it's not the same as
the speed that the thing talks on the wire.

On wireless, the data is in bursts @ say 54Mbps signalling rate although
there are large periods of silence (relatively speaking) so throughput
is lower.

Add to this the fact that wireless with a single radio is half duplex so
that cuts down your throughput then factor in protocol overhead and it's
no wonder that you don't get 54Mbps throughput even though the radio is
chirping away sending data in 54Mbps bursts.

David.

David,

Thanks for the in depth description. I think I understand the
difference betw. signaling rate and throughput rate.

Now the question is, if a wireless G adapter says that it "signals" at
108 Mbps, does that mean that the throughput will be higher than with a
54 Mbps adapter? I know neither one will have the 108 throughput, but
just wondering if the higher signaling rate on the 108 adapter will
result in higher throughput. Or is it limited by the access point I am
using? Not sure if access points have a Mbps signaling rate or if they
just conform to a wireless standard.

Thanks,

Seth

> Now the question is, if a wireless G adapter says that it "signals" at
> 108 Mbps, does that mean that the throughput will be higher than with a
> 54 Mbps adapter? I know neither one will have the 108 throughput, but

It should provide for faster connection but there are other variables.

> using? Not sure if access points have a Mbps signaling rate or if they
> just conform to a wireless standard.

Doesn't matter whether it's an AP or a PC adapter, the rate at which
bits are transmitted has to conform to expected standards. The
throughput is something else. No reason why you couldn't have a very
fast transmission rate yet couple that to a crap processor, poor buffer
and lousy encryption engine such that the bottleneck isn't the wireless
transmision rate but instead just sends very fast bursts @ whatever rate
but with long gaps in between which kills the throughput.

Bear in mind that i'm providing example text, not figures.

David.

So basically, without knowing the other variables, I really can't know
if the 108 will give me more throughput, right?

I think I am going to get the 54 Mbps adapter, just for practical
purposes. Tivo's website lists that model as "compatible", so if I
have any issues with the setup or connections, I'd rather be able to
say, "hey, I am conforming to your standards".

But this is an interesting topic for me. By the way, where did you
learn about all of this?

Thanks!

Seth

> have any issues with the setup or connections, I'd rather be able to
> say, "hey, I am conforming to your standards".

Pretty much, that's what standards are supposed to be for.

> But this is an interesting topic for me. By the way, where did you
> learn about all of this?

Originally? I did an electronic engineering degree at University.

Although very brief, this is a pretty similar thing but discussing baud
rate and telephone transmission although it still comes in to wireless,
http://en.wikipedia.org/wiki/Baud in that originally a symbol state
transition represented one bit and so up to 600 baud, data rate was 600
bps. Then they started messing with phase too so you got to Quadrature
Amplitude Modulation where 16 bits are transmitted per symbol state
http://www.hait.ac.il/staff/commEng/.../qam/demo.html

Wireless RF is just another set of modulation methods and this is what I
mean about the standards being important, in order to be decoded, you
have to have both ends using the same signalling scheme. They can fall
back and forwards to different modulation methods, signalling rates but
at the end of the day, how often those bursts of data are in the medium
is what you're interested in if you want to measure throughput.

Like I said, it's just like if you could read out loud the whole of a
500 page book in 5 seconds. Only useful if someone else can listen at
the same speed and not much use as far as throughput goes if you could
only do it once a week and didn't have to rest before doing it again.

David.