Frank,
The version of Plan D which is shown will not work due to non-degenerate FWM. It may be possible to find a non-uniform channel plan that works but it will take a little time to prove. The first plot below is a first cut that seems to work. I don't think it will be possible to find a working plan that straddles the zero dispersion window if one of the channels is uncooled due to the wide range of FWM frequencies (see 2nd plot).
For the purposes of illustrating coexistence in the liaison, we may want to show the non-uniform plan below or a simpler LAN-WDM Plan D with all four channels in O+ band.
Non-uniform plan: 1270, 1295, 1326, 1340nm with 3nm passband
Non-uniform plan with uncooled 1270nm
All channels in O+ band (no FWM possible): CS = 1.2THz, PB = 3nm
Regards,
John