A Weight-Optimized BP Decoding Algorithm for JSCC based Faster-than-Nyquist

Faster-than-Nyquist (FTN) transmission is a technology that enhances system capacity and spectral efficiency without the need for additional bandwidth or antennas, thus opening up new possibilities for the future development of mobile communications. However, since the transmission rate exceeds the Nyquist rate, the orthogonality between signals is disrupted, inevitably leading to inter-symbol interference (ISI). This paper proposes a joint source-channel coding (JSCC) based FTN system, where JSCC further exploits the residual redundancy left after source coding to reduce the effects of ISI. In addition, we propose a weight-optimized belief propagation (WOBP) algorithm, which aims to classify the unreliability of variable nodes based on the impact of ISI on the channel codes during decoding, and further mitigate the effects of ISI through weight optimization. Simulation results show that this algorithm improves the performance in the high signal-to-noise ratio (SNR) region compared with the traditional <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$B P$</tex> decoding algorithm, and the system bit error rate (BER) improves by 0.42 dB and 0.62 dB at <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$1 \times 10^{-5}$</tex> using AR3A and AR4JA coding, respectively.