PAM-5 at 1.25 Gbaud
I have received some private emails saying that
I might be underestimating the performance of
my own proposal ...
May be that I am to blame for this in the way
I constructed the spreadsheet for the installed
In the entry corresponding to the "electrical
signal power @ Rx", the PAM-5 at 1.25 Gbaud
proposal would appear to have much less signal
power compared to the other on-off n-WDM
In practice, the input power at the Rx of my proposal
does not differ from the input power of the other
on-off n-WDM proposals. What happens is that,
for the purposes of calculating the SNR, what is
important is the *difference* between adjacent levels.
For a 5-level approach this difference is much
smaller. And it is this difference between adjacent
levels that appears under the entry "electrical
signal power @ Rx".
Of course, this smaller "differential power"
between levels is compensated by the smaller
noise power due to use of a smaller receiver
Jaime E. Kardontchik
San Jose, CA 95131