An Efficient Puncturing Scheme for Polar Codes Based on Fixed Information Set

For traditional polar codes, the code length is limited to a power of two, due to the use of the 2×2 kernel matrix. To obtain flexible code lengths, puncturing is widely considered for polar codes. In this paper, we propose an efficient puncturing scheme based on fixed information set (FIS) of the mother code. First, we estimate the qualities of bit channels using the Gaussian approximation after puncturing. In addition, the search scope of punctured bits in the codeword is significantly reduced to the number of frozen bits. Furthermore, a computationally efficient method for selecting the most promising puncturing pattern is employed. Moreover, a low-complexity recursive method is leveraged to construct the puncturing pattern. Simulation results demonstrate that the proposed puncturing scheme outperforms the state-of-the-art FIS-based puncturing schemes remarkably in terms of error-correction performance.