Early selection speeds up polar decoding with fewer computations
Early Selection of Polar Ensemble Decoding with Successive Cancellation Style Component Decoder
Information Theory
Summary
Decoding error-correcting codes often tries many different guesses to improve accuracy, but this takes a lot of computing power. The authors found a way to pick the most promising guesses early on by checking partial results during decoding. This means fewer guesses need to be fully processed, saving time and energy without losing accuracy. They also designed smart rules, including one using a neural network, to decide which guesses to keep. Their tests show that their method matches full guessing accuracy with less work.
What this means in practice
- •For wireless communication engineers: Implement faster decoding in communication devices by reducing computations while maintaining error correction reliability.
- •For hardware design teams: Design energy-efficient error correction modules by employing early stopping in polar decoding processes.
Authors
Pin-Jing Li, Yu-Chih Huang
Abstract
Parallel ensembles of successive cancellation (SC)-style decoders improve error-rate performance by decoding multiple representations of the received vector. Conventional ensembles, however, keep every member active until completion. To reduce computational cost while retaining the diversity gain of ensemble decoding, we propose early ensemble selection (EES), which ranks members using intermediate decoding states and retains a prescribed numbers of members at each predefined selection stage. Within the EES framework, we propose two selection rules, namely path-metric-based selection (PM-EES) and neural-network-based selection (NN-EES). We analytically characterize the computational cost of EES through stage-dependent decoder log-likelihood ratio (LLR) update counts and provide an information-theoretically motivated criterion for choosing selection stages. Simulations on the \((128,64)\) extended BCH code and the \(\operatorname{RM}(4,8)\) code show that suitable selection schedules approach full-ensemble error-rate performance with fewer decoder LLR updates.