위성 및 지상파 방송시스템에서 효율적인 FTN 복호 방식 연구

Abstract

In next generation satellite broadcasting system, requirement of high throughput efficiency has been increasing continuously. To increase throughput efficiency and improve bit error performance, FTN (Faster Than Nyquist) method and LDPC codes are employed in new sattelite standard, DVB-S3 (Digital Video Broadcasting - Satellite - Third Generation) system.

This thesis considered three kinds of methods for increase throughput efficiency. Firstly, as conventional one, high coding rate parity matrix in LDPC encoder is considered. Secondly, punctured coding scheme which delete the coding symbol according to appropriate rules is considered. Lastly, FTN method which transmit fater than Nyquist rate is considered. Among of three kinds of methods, FTN method is most efficient in aspect to performance while maintain same throughput efficiency.

Therefore this thesis focus on FTN method to improve performance and throughput efficiency. In FTN method, it obtain successful performance improve by minimizing ISI ( Inter Symbol Interference). To minimize interference, we analyze conventional FTN signal decoding method and propose a decoding scheme to have optimal bit error performance. There are two methods to improve performance of conventional FTN scheme. First, it is the SIC (Successive Interference Canceller) method that removes the FTN interference by obtaining it from subtracting the LDPC decoded bit stream at the receiving end. Decoded signals through remapping them using the FTN mapper and again subtracting interference from the received signal. It is hard to get the accurate amount of interference at the stage of remapping. Second, when a signal has much interference, BCJR (Bahl, Cocke, Jelinek and Raviv) equalization method, which removes the interference firstly and then passes through the LDPC decoder demonstrates a better performance. However, it has a disadvantage that, with iterative decoding using the conventional BCJR equalization scheme, its performance does not improve significantly due to inefficient iteration, applying the same extrinsic input that is applied as the input value for equalization to the LDPC decoding signal.

In this thesis, to improve the efficiency of the iterative decoding scheme of conventional FTN decoding, it propose a scheme, which separates the LDPC decoding signals and applies them as extrinsic input, and investigate the performance of the iterative decoding scheme proposed in this paper by comparing it with the conventional scheme in a simulation according to the increase of the throughput. Using these results, the proposed scheme that it updates extinsic input value closer to the raw signal is more efficient than two conventional schemes.