The research of probabilistic shaping signal transmission scheme based on neural network LLR calculation

In order to improve the decoding accuracy and the BER performance of probabilistic shaping optical fiber transmission system, a scheme of probabilistic shaping signal transmission utilizing neural network based LLR calculation is proposed in this paper. Compared with the traditional transmission scheme based on the maximum logarithmic approximation LLR (ALLR) calculation, the mean square error ratio of LLR between the proposed transmission scheme and the scheme based on the exact LLR (ELLR) is reduced by about 100 times. And the BER of the system is improved by at least 0.1dB. In addition, the computational complexity of the proposed scheme is significantly lower than that of the transmission scheme based on the exact LLR computation method (ELLR).