| With the development of communication technology and increasing of communicationtraffic, high-speed satellite communications which exploit adaptive coding and modulationtechniques are becoming a hot topic of research. Such systems generally combine advancederror correction codes, such as SCCC, LDPC with high order modulation scheme, such asQAM and APSK to achieve high-speed transmission. SNR estimation algorithm and softdemapping algorithm play an important role in the initialization of decoding, the adaptivemode selection and many other aspects.Therefore, studying the performance andcomplexity of these two algorithms in the case of high order modulation, especially APSK,is of great importance.In first part of the paper, the moment-based SNR estimation algorithm is studied. It hasbeen shown that the performance of original moment-based algorithm degradessignificantly with the increasing of SNR in the case of non-constant modulusconstellations.In order to solve this problem, a hybrid algorithm based on partitioning ofconatellation is proposed.The simulation results indicate that the new algorithm caneffectively extend the range of SNR estimation, while it has the same complexity asoriginal algorithm.Secondly, the SNR estimation based on EM (Expectation Maximization) algorithm isresearched. An envelop-based algorithm is proposed and applied to APSK.The simulationresults show the algorithm has a good performance and low complexity, which is valuablein practice.For soft demapping, the high complexity of the classical algorithms is discussedin the paper and then, three low-complexity soft demapping algorithms for QPSK,8PSKaand16APSK are proposed respectively.The BER of SCCC system which exploits the newalgorithms is evaluated by simulation.Finally, the RTL implementation and debugging of SNR estimation algorithm and softdemapping algorithm is discussed. The module design and resource optimization processare introduced according to the specific requirement. |
PDF Full Text Request