Disadvantages of *extreme* MPSK?

Hi:

In the hypothetical situation I describe, dial-up modem 
telecommunications technology devices use extreme MPSK [Multiple Phase 
Shift Keying] as the modulation scheme. This too-good-to-be-true MPSK 
uses only 1-symbol-per-second but with a Graham's-number amount of 
bits-per-symbol.

http://en.wikipedia.org/wiki/Graham's_number

Now that is a lot of bits-per-baud!

If it were possible for such MPSK to exist, what would be the 
disadvantages of it?

I am aware that since its not possible to exist, its a moot point in 
discussing any advantages or disadvantages. However, I am still curious 
as to what the drawbacks would be if it was possible to exist.

One can say the limited data resolution available to both older, analog 
phone lines and contemporary, digital phone lines sets a hard, hard 
limit on the available *bandwidth* for anything going over a phone line.

Then, in response to such a statement [i.e. about bandwidth], I must ask 
my $$$$$$Graham's-# question:

Wouldn't a baud of 1-symbol-per-second use only 1 Hz of bandwidth 
regardless of the amount of bits-per-symbol?


Thanks,

Radium

"Green Xenon [Radium]" <glucegen1@excite.com> wrote in message 
news:48478a31$0$5717$4c368faf@roadrunner.com...
> Hi:
>
> Then, in response to such a statement [i.e. about bandwidth], 
> I must ask my $$$$$$Graham's-# question:
>
> Wouldn't a baud of 1-symbol-per-second use only 1 Hz of 
> bandwidth regardless of the amount of bits-per-symbol?

Wouldn't the required bandwidth depend on the shape or other 
time-domain characteristics of the signalling element?

Rephrased as, "Can a 1-baud signal be designed to use only 1 Hz 
of bandwidth?", I think the answer depends on the noise that 
you have to overcome and the number of bits you stuff into a 
signalling element.  It seems to me that Shannon has something 
to say about this.

"Green Xenon [Radium]" <glucegen1@excite.com> writes:

> Hi:
>
> In the hypothetical situation I describe, dial-up modem
> telecommunications technology devices use extreme MPSK [Multiple Phase
> Shift Keying] as the modulation scheme. This too-good-to-be-true MPSK
> uses only 1-symbol-per-second but with a Graham's-number amount of
> bits-per-symbol.
>
> http://en.wikipedia.org/wiki/Graham's_number
>
> Now that is a lot of bits-per-baud!
>
> If it were possible for such MPSK to exist, what would be the
> disadvantages of it?

You'd have to have a very high SNR to get the equivalent bit error rate
performance of a lower complexity, less spectrally efficient modulation.

That means that you'd either a) have very low noise, or b) very high
signal power. In practice, there are limitations on each of these.

> Wouldn't a baud of 1-symbol-per-second use only 1 Hz of bandwidth
> regardless of the amount of bits-per-symbol?

Yes. But you're ignoring the other key parameter in any such system: SNR
(or Eb/No).
-- 
%  Randy Yates                  % "She has an IQ of 1001, she has a jumpsuit
%% Fuquay-Varina, NC            %            on, and she's also a telephone."
%%% 919-577-9882                % 
%%%% <yates@ieee.org>           %        'Yours Truly, 2095', *Time*, ELO   
http://www.digitalsignallabs.com

Green Xenon [Radium] wrote:
> Hi:
> 
> In the hypothetical situation I describe, dial-up modem 
> telecommunications technology devices use extreme MPSK [Multiple Phase 
> Shift Keying] as the modulation scheme. This too-good-to-be-true MPSK 
> uses only 1-symbol-per-second but with a Graham's-number amount of 
> bits-per-symbol.
> 
> http://en.wikipedia.org/wiki/Graham's_number
> 
> Now that is a lot of bits-per-baud!
> 
> If it were possible for such MPSK to exist, what would be the 
> disadvantages of it?
> 
> I am aware that since its not possible to exist, its a moot point in 
> discussing any advantages or disadvantages. However, I am still curious 
> as to what the drawbacks would be if it was possible to exist.
> 
> One can say the limited data resolution available to both older, analog 
> phone lines and contemporary, digital phone lines sets a hard, hard 
> limit on the available *bandwidth* for anything going over a phone line.
> 
> Then, in response to such a statement [i.e. about bandwidth], I must ask 
> my $$$$$$Graham's-# question:
> 
> Wouldn't a baud of 1-symbol-per-second use only 1 Hz of bandwidth 
> regardless of the amount of bits-per-symbol?
> 
> 
> Thanks,
> 
> Radium

My guess is the EVM requirements might just be a bit strict.

On Wed, 04 Jun 2008 23:39:45 -0700, "Green Xenon [Radium]"
<glucegen1@excite.com> wrote:

>Hi:
>
>In the hypothetical situation I describe, dial-up modem 
>telecommunications technology devices use extreme MPSK [Multiple Phase 
>Shift Keying] as the modulation scheme. This too-good-to-be-true MPSK 
>uses only 1-symbol-per-second but with a Graham's-number amount of 
>bits-per-symbol.
>
>http://en.wikipedia.org/wiki/Graham's_number
>
>Now that is a lot of bits-per-baud!
>
>If it were possible for such MPSK to exist, what would be the 
>disadvantages of it?

If you look at a bandwidth-efficiency plane showing the capacity of
various modulation types, QAM has higher capacity than PSK after
8-PSK.   16-QAM has a higher capacity than 16-PSK, and the difference
gets bigger as the order increases.

This is not difficult to understand intuitively since with PSK the
distance between nearest neighbor constellation points gets smaller
much faster than the average distance between all points, while with
QAM the available space is being used more completely.

So if one wanted to stretch the conceptual extremes, one is better off
doing it with QAM than PSK, anyway.

>I am aware that since its not possible to exist, its a moot point in 
>discussing any advantages or disadvantages. However, I am still curious 
>as to what the drawbacks would be if it was possible to exist.
>
>One can say the limited data resolution available to both older, analog 
>phone lines and contemporary, digital phone lines sets a hard, hard 
>limit on the available *bandwidth* for anything going over a phone line.
>
>Then, in response to such a statement [i.e. about bandwidth], I must ask 
>my $$$$$$Graham's-# question:
>
>Wouldn't a baud of 1-symbol-per-second use only 1 Hz of bandwidth 
>regardless of the amount of bits-per-symbol?

Pretty much.

As others have pointed out, though, the required EVM and SNR get
impractical at some point.


Eric Jacobsen
Minister of Algorithms
Abineau Communications
http://www.ericjacobsen.org

Blog: http://www.dsprelated.com/blogs-1/hf/Eric_Jacobsen.php

Green Xenon [Radium] wrote:
> Hi:
> 
> In the hypothetical situation I describe, dial-up modem 
> telecommunications technology devices use extreme MPSK [Multiple Phase 
> Shift Keying] as the modulation scheme. This too-good-to-be-true MPSK 
> uses only 1-symbol-per-second but with a Graham's-number amount of 
> bits-per-symbol.
> 
> http://en.wikipedia.org/wiki/Graham's_number
> 
> Now that is a lot of bits-per-baud!
> 
> If it were possible for such MPSK to exist, what would be the 
> disadvantages of it?
> 
> I am aware that since its not possible to exist, its a moot point in 
> discussing any advantages or disadvantages. However, I am still curious 
> as to what the drawbacks would be if it was possible to exist.
> 
> One can say the limited data resolution available to both older, analog 
> phone lines and contemporary, digital phone lines sets a hard, hard 
> limit on the available *bandwidth* for anything going over a phone line.
> 
> Then, in response to such a statement [i.e. about bandwidth], I must ask 
> my $$$$$$Graham's-# question:
> 
> Wouldn't a baud of 1-symbol-per-second use only 1 Hz of bandwidth 
> regardless of the amount of bits-per-symbol?

Such an approach is extreme overkill. All that's really needed is a 
single pulse. The receiver measures its amplitude exactly as a binary 
number with teraabit precision. Then the bits are taken in groups of 
eight and construed as ASCII. The whole of the Library of Congress can 
be encoded in a single pulse.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

On Jun 5, 10:29 am, Eric Jacobsen <eric.jacob...@ieee.org> wrote:
> On Wed, 04 Jun 2008 23:39:45 -0700, "Green Xenon [Radium]"
>
[snip]
>
> >Wouldn't a baud of 1-symbol-per-second use only 1 Hz of bandwidth
> >regardless of the amount of bits-per-symbol?
>
> Pretty much.
>
> As others have pointed out, though, the required EVM and SNR get
> impractical at some point.
>
> Eric Jacobsen

To add to Eric's reply, don't forget that Eb/No for most systems is
limited by component characteristic.  So your 1024-PSK will
probably require extreme specs in terms of DAC/ADC quality,
timing and clock synchronization, amplifier quality, etc.

Julius

Jerry Avins <jya@ieee.org> writes:
> [...]
> Such an approach is extreme overkill. All that's really needed is a
> single pulse. The receiver measures its amplitude exactly as a binary
> number with teraabit precision. Then the bits are taken in groups of
> eight and construed as ASCII. The whole of the Library of Congress can
> be encoded in a single pulse.

Now why didn't I think of that? :)
-- 
%  Randy Yates                  % "She's sweet on Wagner-I think she'd die for Beethoven.
%% Fuquay-Varina, NC            %  She love the way Puccini lays down a tune, and
%%% 919-577-9882                %  Verdi's always creepin' from her room." 
%%%% <yates@ieee.org>           % "Rockaria", *A New World Record*, ELO   
http://www.digitalsignallabs.com

Randy Yates wrote:


> "Green Xenon [Radium]" <glucegen1@excite.com> writes:
> 
>> Hi:
>>
>> In the hypothetical situation I describe, dial-up modem
>> telecommunications technology devices use extreme MPSK [Multiple Phase
>> Shift Keying] as the modulation scheme. This too-good-to-be-true MPSK
>> uses only 1-symbol-per-second but with a Graham's-number amount of
>> bits-per-symbol.
>>
>> http://en.wikipedia.org/wiki/Graham's_number
>>
>> Now that is a lot of bits-per-baud!
>>
>> If it were possible for such MPSK to exist, what would be the
>> disadvantages of it?


> 
> You'd have to have a very high SNR to get the equivalent bit error rate
> performance of a lower complexity, less spectrally efficient modulation.
> 
> That means that you'd either a) have very low noise, or b) very high
> signal power. In practice, there are limitations on each of these.


Doesn't MPSK require less of a dynamic range than ASK? MPSK's signals 
have a constant peak-to-peak amplitude while ASK's PtPA varies according 
to the 1s and 0s.


> 
>> Wouldn't a baud of 1-symbol-per-second use only 1 Hz of bandwidth
>> regardless of the amount of bits-per-symbol?


> 
> Yes. But you're ignoring the other key parameter in any such system: SNR
> (or Eb/No).


What's Eb/No?

Karl Molnar wrote:


> Green Xenon [Radium] wrote:
>> Hi:
>>
>> In the hypothetical situation I describe, dial-up modem 
>> telecommunications technology devices use extreme MPSK [Multiple Phase 
>> Shift Keying] as the modulation scheme. This too-good-to-be-true MPSK 
>> uses only 1-symbol-per-second but with a Graham's-number amount of 
>> bits-per-symbol.
>>
>> http://en.wikipedia.org/wiki/Graham's_number
>>
>> Now that is a lot of bits-per-baud!
>>
>> If it were possible for such MPSK to exist, what would be the 
>> disadvantages of it?
>>
>> I am aware that since its not possible to exist, its a moot point in 
>> discussing any advantages or disadvantages. However, I am still 
>> curious as to what the drawbacks would be if it was possible to exist.
>>
>> One can say the limited data resolution available to both older, 
>> analog phone lines and contemporary, digital phone lines sets a hard, 
>> hard limit on the available *bandwidth* for anything going over a 
>> phone line.
>>
>> Then, in response to such a statement [i.e. about bandwidth], I must 
>> ask my $$$$$$Graham's-# question:
>>
>> Wouldn't a baud of 1-symbol-per-second use only 1 Hz of bandwidth 
>> regardless of the amount of bits-per-symbol?
>>
>>
>> Thanks,
>>
>> Radium


> 
> My guess is the EVM requirements might just be a bit strict.
> 


Does EVM stand for "Error Vector Magnitude"?