Improved CA-SCL Decoding of Polar Codes Based on Perturbation

This paper mainly discusses how to improve decoding performance of polar codes under imperfect channel estimation. By adding additional appropriate perturbation value to the received log likelihood ratio (LLR), it can effectively resist the negative effects of imperfect channel estimation. The decoding performance and computing complexity of the newly proposed method have been compared to conventional methods under different SNR conditions, and results show that with only a slight increase of decoding time, the decoding performance has been improved about 0.1dB.