Low Complexity Lossy Compression Using Polar Codes

When the code length tends to infinity, polar codes have been proved to achieve the rate-distortion bounds of binary and Gaussian sources. However, the ideal conditions cannot be achieved in reality. In this paper, we propose a source coding scheme for binary sources based on polar codes, which has good performance for short-codes. As for Gaussian sources, a low-complexity source coding algorithm is proposed. In this scheme, we combine the characteristics of source polar coding with variable code rate and design a quantization algorithm suitable for Gaussian sources, which makes the system has lower complexity compared to the existing continuous source lossy compression system, and maintains good compression performance.