5. Nonlinearity Mitigation in Coherent Optical Communication Systems: All-Optical and Digital Signal Processing Approaches

Information transmission through fiber-optic channel is subjected to several impairments such as chromatic and polarization mode dispersion, nonlinear phase noise due to interaction of amplifier noise with fiber Kerr nonlinearity, and nonlinear effects such as self- and cross-phase modulation. In addition, laser phase noise and frequency offset between signal and local oscillators also degrade the received signal quality. Unless these impairments are mitigated, the performance of high data rate optical communication systems is degraded. We describe two approaches to mitigate fiber impairments in high data rate coherent optical communication systems. In the first approach, nonlinearity and dispersion in either fibers or semiconductors is used to undo the effects of transmission fiber on the optical carrier. We propose the use of mid-span spectral inversion, realized using counter-propagating dual pumped four-wave mixing in fibers, to mitigate dispersion and nonlinearity in 40 Gbps QPSK systems. We describe our work on realizing optical phase conjugation in semiconductor optical amplifiers. In the second approach, the optical signal is sampled after coherent reception and processed using digital signal processing algorithms to mitigate dispersion and nonlinearity. We describe Kalman filters to estimate and track phase noise in 100 Gbps QPSK systems. We also describe radial basis function neural network equalizer to mitigate nonlinearity in 80 Gbps 16 QAM CO-OFDM systems.