All-optical nonlinearity mitigation in fiber-optic communications

Abstract: The two main factors limiting the data throughput in modern fiber-optic communication links are the noise added by amplifiers and the nonlinear response of the optical fiber due to the Kerr effect. Today there are communication systems operating remarkably close to the limits set by these two factors. In order to increase the data throughput in a single fiber one can attempt reducing the noise added by the amplifiers or mitigate the nonlinear distortion which is dominated by deterministic effects. The work presented in this thesis is focused on reducing the negative impact of the Kerr nonlinearity through the use of all-optical signal processing by transmission of a phase-conjugated copy alongside the signal. A concept where the phase-conjugated data is repeated in time domain is investigated and it was found that it performs comparably to the conventional phase-conjugated twin waves concept. The rest of the thesis is dedicated to studying various aspects of the two-mode copier-phase-sensitive amplifier (PSA) scheme. These studies focus both on improving the understanding and on optimizing the performance of the nonlinearity mitigation in copier-PSA links. We study the optimization of the dispersion map on a single- and two-span basis and it is shown in simulations that significant improvements can be achieved by optimization over two spans. In a numerical study it was found that there is potential for improving the efficiency of the nonlinearity mitigation by addition of dsitributed Raman amplification (DRA) in a copier-PSA link. Long-haul transmission using the copier-PSA scheme is demonstrated experimentally both with and without DRA. Without DRA at 10 Gbaud, it is shown that it is possible to improve the transmission reach by up to a factor of 5.6 with the addition of PSA. With DRA at 28 Gbaud, the improvement in transmission reach is smaller but we observe an increase in the optimum launch power when enabling the PSA indicating improved nonlinearity mitigation.