Online recursive independent component analysis based equalization for orbital angular momentum mode division multiplexed transmission

In this paper, we propose an online recursive independent component analysis (ORICA) algorithm as an alternative to multi-input multi-output (MIMO) equalizers in orbital angular momentum (OAM) mode division multiplexed (MDM) transmission systems. Due to the random intra-group mode coupling in the OAM system and the application limitations of the algorithm itself, the traditional MIMO-based constant modulus algorithm (CMA) algorithm is not suitable for OAM transmission. To address this issue, we introduce an adaptive forgetting factor into the conventional independent component analysis (ICA) algorithm and iterate the whiting matrix and weigh matrix. As a result, the whole algorithm can achieve efficient signal equalization and demodulation. A 10 km ring core fiber (RCF) transmission system, which transfers eight different OAM multi-mode QPSK signals, was set up to verify the effectiveness and adaptability of the algorithm. Our experimental results show that the proposed ICA algorithm outperformed CMA with an average improvement in bit error rate (BER) of 0.6 dB. In addition, the proposed ICA algorithm outperformed other conventional ICA algorithms with an average improvement in BER of 9.28 dB. Compared to CMA and Infomax ICA, the complexity of the ORICA algorithm is reduced by 42.6% and 29.2%, respectively, demonstrating that this algorithm achieves high-precision demodulation while maintaining low complexity.