Progressive Freezing Low-Complexity Belief Propagation Decoder of Polar Codes for 6G Wireless Communications

Abstract

The high computational complexity of belief propagation (BP) for polar codes, which results in low energy efficiency, constrains the application of polar codes in terabits-per-second throughput scenarios of future sixth-generation (6G) wireless communication systems. In this paper, we propose a progressive freezing (PF) criterion for the BP decoder of polar codes to avoid redundant computational complexity. First, we have demonstrated that an upper bound exists for the absolute value of each soft messages based on the update rule of the BP decoder. Then, we propose the PF criterion, in which the computation in subsequent iterations can be avoided once the absolute value of soft message reaches its upper bound. Simulation results indicate that the PF criterion can circumvent 22.56% of redundant computations and achieve a performance gain of 0.1 dB in low signal-to-noise ratio regions, which facilitates the application of polar codes in 6G high-throughput scenarios.