Recurrent neural nets achieving MLSE performance in bandlimited optical channels

We explore long short-term memory-recurrent neural networks (LSTM-RNN) for mitigating severe inter-symbol-interference caused by bandlimited components. We simulate two bandlimited channels (experimental and synthetic) and demonstrate our proposed LSTM-RNN architecture provides the same bit error rates as a maximum likelihood sequence estimator (MLSE) for 4QAM.