A time sharing (TS) hybrid probabilistic shaping (HPS) scheme for nonbinary (NB) LDPC codes is proposed. For the conventional probabilistic shaping (CPS) scheme, the demodulator uses bit-metric decoding (BMD) for binary LDPC codes and NB LDPC codes. However, the demodulator for NB LDPC codes adopts symbol-metric decoding (SMD) on the AWGN channel to avoid the loss of symbol-to-bit mapping and obtain better performance than binary LDPC codes. In order to take advantage of SMD demodulation for NB LDPC codes, the NB-LDPC information symbols use constant composition distribution matcher (CCDM) to generate quadrature amplitude modulation (QAM) symbols with unequal probability distribution, while the NB-LDPC parity symbols are utilized to yield QAM symbols with approximately uniform distribution. The SMD demodulation of the proposed NB-LDPC-coded TS HPS scheme has lower demodulation complexity than the NB-LDPC-coded CPS scheme. The information-theoretical analysis of the CPS system and the TS HPS system turns out that the proposed NB-LDPC-coded TS HPS scheme is superior to the NB-LDPC-coded CPS scheme for 8QAM and 32QAM format. Simulation results verify the theoretical average mutual information (AMI) analysis and show that the proposed NB-LDPC-coded TS HPS scheme has up-to-0.24 dB and 1.10 dB performance gain compared with the NB-LDPC-coded CPS scheme for 8QAM and 32QAM format, respectively. Compared with the binary LDPC-coded regular Nyquist scheme for 8QAM and 32QAM format, the proposed NB-LDPC-coded TS HPS scheme has up-to-1.31 dB and 2.04 dB performance gain, respectively. Meanwhile, the LDPC code rate of the proposed TS HPS scheme is more flexible than the CPS scheme. Therefore, the proposed NB-LDPC-coded TS HPS scheme is more reliable, efficient, and flexible for 8QAM and 32QAM, and thus suitable for the sixth generation (6G) communication system.INDEX TERMS Time sharing hybrid probabilistic shaping (HPS), bit-metric decoding (BMD), symbol-metric decoding (SMD), nonbinary (NB) LDPC, average mutual information (AMI).
I. INTRODUCTION
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.