Implementation And Application Of A Rateless Coded Modulation Method

In order to meet the growing demand for high-speed data transmission and improve the utilization ratio of the channel, this thesis suggests a rate compatible modulation(RCM) which is able to approach channel capacity. Although channel information feedback is not necessary for RCM, smooth rate adaptive and seamless channel matching can be achieved. RCM also has the rateless characteristics, which is similar to the traditional digital fountain code. As long as the receiver receives enough symbols or data packets, the decoder is able to successfully decode. In order to improve the traditional RCM’s performance, this thesis suggests two methods of constructing the generation matrix of RCM. The improved schemes can be applied to higher dynamic range of SNR and produce higher throughput. In this thesis, a low-complexity decoding algorithm is proposed, which can greatly reduce the complexity of decoding in the case of a small rate loss. In the end, the application of RCM is studied.First of all, this thesis analyzes the basic principles of RCM, as well as the adaptive coded modulation method based on RCM. Secondly, according to the principle and characteristics of RCM, the basic criteria of designing RCM generating matrix are studied. Aimed at eliminating the shortcomings of RCM generation matrix, this thesis presents two designing methods for RCM generation matrix by optimizing the weight distribution and degree distribution of generation matrix. A stop criterion for iterative decoding is also proposed. The theoretical analysis and experimental simulations prove that the new method can obtain better performance. Simulation results show that whether in the high signal-to-noise ratio or low signal-to-noise ratio, the new coded modulation method of LDPC codes combined with RCM is better than the adaptive systems of LDPC codes, turbo codes and Raptor codes combined with QAM modulation. Finally, to make full use of the characteristic that the nonzero elements in the RCM generator matrix G are real numbers, a low-complexity decoding method is proposed. This method can reduce the complexity of horizontal iteration by about a half, greatly reducing the complexity of decoding.Since RCM has no need for channel information feedback, has the characteristic of rate adaptive, and can use LDPC code or other error correction code for pre-encoding, RCM has a great potential. RCM may be an alternative to satellite communications in the future. This thesis presents a secure encoding program for RCM, and the application of RCM in radio communication is introduced.