Theory And Hardware Research Of Novel Shaping Coded Modulation Based On Faster-Than-Nyquist Signaling

Presently in communication systems, with the data skyrocketing, the limitation of bandwidth is becoming critical. Faster-than-Nyquist (FTN)signaling has huge potential because it can increase the spectral efficiency.Based on this reason, this thesis accomplishes profound research around FTN signaling.In terms of algorithm, this thesis presents basic information of FTN signaling, especially the operations inside FTN transmitter and FTN receiver. An iterative scheme between equalization and decoding based on Turbo architecture is designed to compensate the internal Inter-Symbol Interference (ISI). Furthermore, this thesis reduces the complexity of FTN receiver according to the Ungerboeck model and the symmetry of ISI taps.As for the system design, this thesis compares FTN system with Nyquist system. The results demonstrate that FTN system obtains 1.77dB and 1.4dB error performance gain when the equivalent code rates are 3/4 and 2/3 respectively. In addition, this thesis delves into the influence of different parameters inside FTN system such as non-orthogonal shaping pulses and different compressed factors. FTN system outperforms Nyquist system in all situations.What's more, this thesis implements a FTN transceiver on FPGA platform. Several techniques are employed in this hardware implementation to take advantage of hardware's parallel computing feature,such as Ping-Pong buffer, reusable module, multi-core scheme and pipelining. Consequently, this implementation reduces the time delay and boosts its throughput. The maximum clock frequency is over 200MHz and the deterioration of error performance in the worst case is only 0.2dB compared with that of software simulation.Moreover, this thesis provides knowledge about FTN system on intermediate frequency range. The final test indicates that the FTN transceiver works smoothly corresponding to the proposed algorithms.Thus, the practicality of this thesis is confirmed.