Parallel and Pipelined CMA for High-Speed and Real-Time Optical Coherent Receivers

In order to reduce the logic resources required by parallel polarization demultiplexing algorithm, an improvement scheme based on parallel structure of constant mode algorithm (PCMA) was proposed in this paper. We designed a new error accumulation method in this improvement scheme. After that, we make a simulation of our algorithm at a data rate of 10-Gb/s. The simulation results show that the optimized algorithm can accelerate the convergence speed and reduce the steady-state mean square error (MSE) and bit error rate (BER).