Faster than Nyquist Signaling with Spatial Coupling

Sammanfattning

The emergence of the Faster-Than-Nyquist signaling (FTN) technology has im- proved the tight spectrum resources by compressing the transmission interval between adjacent pulses. In a natural environment, data transmission via a wire- less medium and through multi-path, at a high data rate, the leakage of these paths among symbols can cause severe inter-symbol interference (ISI). In order to compensate for the ISI caused by FTN, turbo equalization technology is demon- strated in this thesis. A challenge when implementing turbo-equalization is the required matching be- tween channel codes and equalizer, leading to a fundamental trade-off between waterfall performance and error floor in different Signal-to-Noise (SNR) regions. Spatially coupled low-density parity-check codes (SC-LDPC codes) are demon- strated and optimized by using window decoder to improve the system perfor- mance. In this thesis, we mainly investigate SC-LDPC codes with FTN signaling and turbo equalization. The simulation is written in MATLAB together with some MEX files. Firstly, FTN signaling with uncoupled convolutional codes and LDPC codes is simulated as a comparison to a previous article, then SC-LDPC codes are also implemented to evaluate the performance improvement according to chang- ing the different parameters, such as code rate R and time-squeezing factor τ. Meanwhile, the spectral efficiency performance simulation with different code rates and time acceleration factors is also discussed. The results show that by im- plementing FTN signaling, turbo equalizer and spatial coupling together in the system, such a combination can achieve better performance than the traditional Nyquist scheme in bit error rate (BER) and spectral efficiency (SE) aspects.