Noise-Aided Belief Propagation List Decoding of Polar Codes

Abstract

Noise can help weak information bearing signal to be detected in nonlinear systems by the use of stochastic perturbation. In this letter, we propose a novel belief propagation list (BPL) decoder that relies on artificial noise. The proposed decoder is constructed by using parallel independent BP decoders. Noise with different powers are added to different BP decoders that runs in parallel, and the output of the decoder that pass the early detection and termination criteria check is considered as the recovered data. Adding small amount of noise enables the decoder to handle un-converged errors. The obtained results show that the performance of noise-aided BPL decoder is much better than the performances of BP, successive cancelation (SC), and successive cancelation list (SCL) decoders, and it is a bit behind of cyclic redundancy check aided SCL in terms of bit-error rate performance.