Santosh,

Usually the advantage of single-dimensional modulation is in
implementation complexity, as you've alluded, as well as some possible
noise immunity assuming one doesn't overdo it (e.g., BPSK).   In
general, though, if one wants to preserve spectral efficiency then
higher-dimensions are usually the way to go.

Other considerations come from the nature of the channel or
transmission medium.  Many modulation applications are at
baseband-only or near-baseband (e.g., single channel wireline,
recording systems, etc.), and for these the single-dimensional
modulations may have an overall system advantage.

As usual, the needs of the entire system need to be considered to
select a modulation type.  Whether you need power efficiency to be the
priority or spectral efficiency or implementation cost/complexity
makes a huge difference in selecting the appropriate modulation type.
There is, IMHO, no "best" overall modulation type.

Hope that helps a bit,

Eric

On 2 Sep 2003 14:05:36 -0700, santosh.nath@ntlworld.com (santosh nath)
wrote:

>Hi All,
>
>I have collected few classic papers on channel coding,MLSE and Viterbi
>Algorithm written by the pioneers during 1966-1972 - Including works
>of Hancock,Gallager,Viterbi,G.D.Froney,Omura,Lucky etc. I have found a
>later paper written by Ungerboeck at IEEE trans on info
>theory,Jan,1982 where he introduced the idea of set partitioning later
>popularized by S U H Quereshi et al as RSSE - a reduced complexity
>version of MLSE.
>
>He has mentioned that there may not be significant difference in gains
>achievable with one or two dimesional coded modulation. As far as I
>remember most practical digital modulations except BPSK are two
>dimensional(e.q QPSK,8PSK,16QAM,32QAM,64QAM etc.)- representation of
>signal constellation.
>
>My question is :
>
>Are there many examples of bandwidth efficient one dimensional coded
>modulation? Where are they applied? For example, he has mentioned
>2-AM,
>4-AM,8-AM,16-AM representing signal points on X axis. Frankly, I am
>not very familiar with these amplitude modulations.
>The advantage here could be simpler multiplication since the phasors
>lies on real X axis and one could avoid complex multiplication during
>processing.
>
>Interested reader can look into the famous paper:
>"Channel coding with multilevel/phase signals",Ungerboeck,IEEE Trans.
>on Info theory,Jan,1982.
>
>Regards,
>Santosh

Eric Jacobsen
Minister of Algorithms, Intel Corp.
My opinions may not be Intel's opinions.
http://www.ericjacobsen.org

Hi All,

I have collected few classic papers on channel coding,MLSE and Viterbi
Algorithm written by the pioneers during 1966-1972 - Including works
of Hancock,Gallager,Viterbi,G.D.Froney,Omura,Lucky etc. I have found a
later paper written by Ungerboeck at IEEE trans on info
theory,Jan,1982 where he introduced the idea of set partitioning later
popularized by S U H Quereshi et al as RSSE - a reduced complexity
version of MLSE.

He has mentioned that there may not be significant difference in gains
achievable with one or two dimesional coded modulation. As far as I
remember most practical digital modulations except BPSK are two
dimensional(e.q QPSK,8PSK,16QAM,32QAM,64QAM etc.)- representation of
signal constellation.

My question is :

Are there many examples of bandwidth efficient one dimensional coded
modulation? Where are they applied? For example, he has mentioned
2-AM,
4-AM,8-AM,16-AM representing signal points on X axis. Frankly, I am
not very familiar with these amplitude modulations.
The advantage here could be simpler multiplication since the phasors
lies on real X axis and one could avoid complex multiplication during
processing.

Interested reader can look into the famous paper:
"Channel coding with multilevel/phase signals",Ungerboeck,IEEE Trans.
on Info theory,Jan,1982.

Regards,
Santosh