[A-2] Information Theory

9:00 AM - 9:15 AM JST(12:00 AM - 12:15 AM UTC)Candidate for Academic Incentive Award

[A-2-02] A Study on Design SNR Optimization of Polar Codes for Successive Cancellation List Decoding

◎Hidetake Matsui Matsui¹, Hideki Ochiai¹ (1. The University of Osaka)

Keywords:

Polar codes,Successive cancellation list decoding,Signal to noise ratio,Weight distribution,Reed-Muller codes

Polar codes can achieve the channel capacity under successive cancellation (SC) decoding as the code length tends to infinity. However, for practical finite code lengths, their block error rate (BLER) under SC decoding is inferior to those of LDPC and turbo codes. Therefore, successive cancellation list (SCL) decoding has been proposed as a more powerful decoding method. Although conventional polar-code construction methods have been developed assuming SC decoding, the resulting constructions are not necessarily effective for SCL decoding. In this paper, we optimize the design signal-to-noise ratio (SNR), a parameter in conventional construction methods, to construct polar codes suitable for SCL decoding. The effectiveness of the proposed approach is demonstrated through computer simulations.