Improved Segmented SC-Flip Decoding of Polar Codes Based on Gaussian Approximation

Compared to Successive Cancellation (SC) decoding algorithm, Successive Cancellation Flip (SCF) decoding algorithm keep the same average computational complexity at high SNR while improving performance. To correct more errors, Partitioned Successive Cancellation Flip (PSCF) decoding algorithm is subsequently proposed with lower computational complexity than SCF decoding algorithm. In this paper, we propose an improved segmented SC-Flip (SSCF) decoding algorithm, which divide the codeword through Gaussian Approximation (GA) of calculating the error probability of each bit channel. Simulation results show that this segmented approach can achieve 48.66% average complexity reduction at SNR of 0.5 dB for (1024, 512) polar code, and better performance compared to PSCF decoding algorithm.