ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술지 Recursive Design of Precoded Polar Codes for SCL Decoding
Cited 9 time in scopus Download 1 time Share share facebook twitter linkedin kakaostory
저자
Vera Miloslavskaya, Branka Vucetic, 이용휘, 박기윤, 박옥선
발행일
202112
출처
IEEE Transactions on Communications, v.69 no.12, pp.7945-7959
ISSN
0090-6778
출판사
IEEE
DOI
https://dx.doi.org/10.1109/TCOMM.2021.3111625
협약과제
21HH1200, 셀룰러 기반 산업 자동화 시스템 구축을 위한 5G 성능 한계 극복 저지연, 고신뢰, 초연결 통합 핵심기술 개발, 신재승
초록
A novel method to recursively construct a set of precoded polar codes of various rates and short-to-moderate lengths is presented. The proposed code design method minimizes the successive cancellation (SC) decoding error probability estimate under three constraints. The first constraint is the minimum distance requirement to improve the maximum-likelihood (ML) performance of the resulting code and therefore the performance under the SC list (SCL) decoding. The other two constraints introduce preselected supercode and subcode, where the supercode ensures fast computation of the minimum distance and the subcode ensures reduction of the search space size. The supercode is given by the Plotkin sum of shorter codes, which are nested to simplify computation of low-weight codewords. These low-weight codewords are needed to satisfy the minimum distance constraint. The simulation results indicate that the proposed precoded polar codes of lengths 128 and 256 provide a better frame error rate (FER) than polar codes with CRC and e-BCH polar subcodes under the SCL decoding algorithm with the list size 8-128.
KSP 제안 키워드
Code design, Decoding error, Design method, Minimum distance, Polar codes, Probability estimate, Reduction of the Search Space Size, Successive cancellation, decoding algorithm, distance constraint, error probability