Back to Search Start Over

Hybrid Coupled Serially Concatenated Codes.

Authors :
Yang, Chaojie
Zhao, Shancheng
Ma, Xiao
Source :
IEEE Transactions on Communications. Jul2022, Vol. 70 Issue 7, p4301-4315. 15p.
Publication Year :
2022

Abstract

In this paper, we present a class of hybrid coupled serially concatenated codes (HC-SCC), in which coupling is achieved by re-encoding partial outputs from both the outer and the inner encoders. Firstly, we derive the exact density evolution (DE) equations over the binary erasure channels (BECs) for the HC-SCCs, which can be used to analyze the impact of the coupling ratios and the coupling memories on the performances of the HC-SCCs. Secondly, we present a genie-aided lower bound for the proposed HC-SCCs to estimate their error-floors. Thirdly, we use the DE equations to design randomly punctured HC-SCCs, including rate-compatible HC-SCCs. The analytical and the numerical results show that: 1) for comparable coupling memories, the iterative decoding thresholds of the proposed HC-SCC ensembles are better than those of the recently introduced spatially coupled serially concatenated code (SC-SCC) ensembles and the partially parity-coupled turbo code (PPC-TC) ensembles; 2) with outer coupling memory five and inner coupling memory five, the iterative decoding thresholds of the rate-compatible HC-SCCs over the BECs are 0.0003 away from the corresponding channel capacities for a wide range of coding rates; 3) simulation results over the additive white Gaussian noise (AWGN) channels and the BECs are provided to confirm the performance advantages of the HC-SCCs. Particularly, the simulation results show that the HC-SCCs admit lower error-floor than the SC-SCCs and the rate-compatible HC-SCCs perform better than the PPC-TCs, which are consistent with the results of DE analysis. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISSN :
00906778
Volume :
70
Issue :
7
Database :
Academic Search Index
Journal :
IEEE Transactions on Communications
Publication Type :
Academic Journal
Accession number :
158023002
Full Text :
https://doi.org/10.1109/TCOMM.2022.3175525