Based On The Continuous Phase Modulation Capacity Analysis Of Cooperative Diversity System

Thanks to the property of constant envelope and continuous phase, the signalsgenerated by Continuous Phase Modulation (CPM) is able to apply low-cost,high-energy-efficiency, nonlinear power amplifier, which is desirable in a cost andenergy limited cooperative communication system.Symmetric information rate, which can be considered as the lower bound ofsystem capacity, in the CPM-based cooperative communication system is studied.Aiming at maximizing ergodic capacity, the optimal power allocation strategy is alsoanalyzed. Furthermore, the effects of SNR estimate error and channel estimate error onsystem capacity has been studied and a novel, robust iterative SNR estimate algorithmis proposed based on the results，and efficiency of the algorithm has also been verifiedvia analysis and simulation.Firstly, the system capacity of CPM-based cooperative system which adopts theBCJR demodulation algorithm is studied, and the analysis and simulation results showthe cooperative CPM system have larger capacity and lower outage probability thanCPM-based SISO system.Secondly, aiming at maximizing ergodic capacity, the optimal power allocationstrategy is derived. When the relay node is near to the source node, power allocationstrategy has little effect on the system capacity. On the contrary, when the relay node isfar from the source node, the source node need more power than the relay node. Andthe system with optimal power allocation strategy can obtain larger capacity than thesystem with equal power allocation strategy.Thirdly, effects of SNR estimate error and channel estimate error on capacity hasbeen studied. Analysis and simulation results show that, BER has little loss but SIR hasa considerable loss when SNR estimate error increase. The system capacity can reachthe peak with increasing SNR if there is a SNR estimate error. With the SNRincreasing, channel estimate error becomes the key factor affecting the system BER,and the BER has a larger loss. The system capacity can’t reach the peak withincreasing SNR if there is a channel estimate error. Finally, iterative SNR estimate algorithm based on Rimoldi decomposition isproposed in the CPM-based cooperative system. Through Rimoldi decomposition,SNR can be iteratively estimated to update the MMSE equalizer, so that a moreaccurate estimated SNR and higher system efficiency can be obtained. The analysisand simulation results show that the system without SNR update needs5dB extrapower compared to the system with the proposed algorithm. And the performance ofthe system with the proposed algorithm is close to the system with perfect channelinformation. Furthermore, the paper studied the effect of frame length on iterativeestimate error of SNR, and the optimal frame length is derived by the analysis andsimulation results.