To my understanding the main difference between colored and non-colored signal is that non-colored in 1310/1550nm while the colored signal is lambda-specific (anywhere in the C-Band or any other band used in DWDM).

Can anybody explain more?

good answer

I agree with you, waiting for other engineer to answer.

In my understanding [forgive for English :)]:
1 . Non-colored interface:
- Modulation: has simple modulation (for example, amplitude [ASK]). As a result, they have the parameters of a spectral form and width.
- Signal structure: has the speed equal to a client signal (for example, 10.31 Gbit/s - 10GE LAN, 9.95 Gbit/s - 10GE WAN, 99.53 Gbit/s - 100GE, 2.488 Gbit/s - STM16, 9.95 Gbit/s - STM64, 39.8 Gbit/s - STM256... ). This speed is less, than at color interfaces.

- Wave length: has one of standard lengths of waves (1310 nanometers, 1550 nanometers... ). Requirements to a deviation of the central frequency the very soft. Therefore, and requirements to the laser too.
2 . Colored interface:
- Modulation: often has more difficult linear modulation (for example, options of difficult amplitude modulation [QAM... ] options of phase modulation [QPSK, BPSK... ]... ).
- Signal structure: has higher linear speed due to addition of various office information and FEC (for example, 11.049 Gbit/s - OTU1e (10GE LAN), 111.809 Gbit/s - OTU4 (100GE), 2.666 Gbit/s - OTU1 (STM16), 10.709 Gbit/s - OTU2 (STM64), 43.018 Gbit/s - OTU3 (STM256)... ).
- Wave length: has one of lengths of the waves, defined in a grid of frequencies of ITU-T (for example, 1547.715 nanometers, 1546.119 nanometers, 1544.526 nanometers... ). These lengths of waves are various for a grid of frequencies 50GHz, 100GHz, 200GHz. Respectively, rigid requirements for a deviation of the central frequency are imposed to them from norm. It doesnt concern superchannels where FlexGrid is applied.

3 . The coding, as a rule, at them identical (for example, NRZ).

Good answer! Thanks!

good answer

Nicely explained. Thanks

Good for sharing, thanks