Feedforward neural network-based chaos encryption method for polarization division multiplexing optical OFDM/OQAM system

• A feedforward neural network-based chaos encryption (FNNCE) metthod is proposed. • A large key space can be created, resisting the exhaustive attack. • The proposed method can enhance the security of PDM O-OFDM/OQAM system. Optical orthogonal frequency division multiplexing offset quadrature amplitude modulation （O-OFDM/OQAM), as a technique that has a great potential to provide high spectral efficiency optical transmission in physical layer for future communication systems, faces challenges in data transmission security like malicious eavesdropping. In this letter, a novel twice encryption method based on chaotic mapping and feedforward neural network (FNN) is proposed to achieve the capability of anti-eavesdropping and security enhancement for polarization division multiplexing (PDM) O-OFDM/OQAM system. In the proposed method, the symbols in an OFDM/OQAM block are first used for symbol substitution by the chaotic sequences generated by Lorenz mapping and then permutated by chaotic scrambling vectors generated by FNN. The simulation results show that the FNN-based chaos encryption (FNNCE) method can strengthen the security of data transmission and realize the large key space simultaneously, and without apparent bit error rate (BER) deterioration in comparison with the original system.