| The distinguishing feature of the high spectrum efficiency of the combination of the Multi-Input Multi-Output(MIMO)system and Filter Bank Multicarrier-Offset Quadrature Amplitude Modulation(FBMC-OQAM)attracts academic attention.However,the high complexity of the MIMO FBMC-OQAM caused by the special structure of the FBMC-OQAM seriously restricts its application.Thus,it’s of realistic significance to design a low-complexity detection algorithm with acceptable performance.This thesis mainly aims to research on signal detection algorithms for MIMO FBMC-OQAM systems.The main contents are outlined as follows.A low complexity MIMO FBMC-OQAM signal detection algorithm based on List Sphere Detection(LSD)algorithm is presented.The performance of the LSD algorithm is closely related with the length of the list.Though a long list brings a low bit error rate(BER),it leads to a high complexity at the same time.This thesis firstly eliminate the inherent interference according to the matrix theory and the numerical characteristics of the MIMO FBMC-OQAM demodulation signal,then a Grover quantum searching method is proposed to utilize the parallelism of quantum to weaken the connection between the length of list and hence calculation amount based on the LSD algorithm is reduced for signal detection.Finally,the simulation results show that the modified LSD algorithm can reduce the complexity of the system while maintaining the same performance as the traditional LSD algorithm.MIMO FBMC-OQAM signal detection algorithm based on BP neural network is proposed.On the basis of eliminating the inherent interference of MIMO FBMC-OQAM signal,the signal of non-interference is set as the input the neural network.Then the signal detection is performed after the weight matrix is close to be stable.In this algorithm,the convergence rate of MIMO FBMC-OQAM signal detection is improved by considering the rule of adaptive learning rate and constructing the connection between error function and learning rate.The simulation results show that the convergence rate is related to the signal-to-noise ratio(SNR)of the MIMO FBMC-OQAM system under certain accuracy of the neural network systems. |
