Interdimensional Interference Equalizing Using Recurrent Neural Network for Multi-Dimensional Optical Transmission

Abstract

During multi-dimensional optical transmission, because of photodetector (PD) operation as a square law detector, signals get distorted when they are modulated on polarization and phase due to interdimensional interference (IDI) by intensity-modulated signals. We propose an IDI mitigation technique using recurrent neural network for multi-dimensional optical transmission. The signal transmission performance of the proposed equalization technique was experimentally analyzed. The performance of the proposed system was verified using the symbol error rate (SER).

Keywords

• Multi-dimensional optical modulation
• interdimensional interference (IDI)
• recurrent neural network (RNN)
• optical access network