Research On Noncoherent Multiuser Detection Technology For Nonorthogonal Multipulse Modulation System

Nonorthogonal multipulse modulation(NMM)technique based on noncoherent detection does not require knowledge of carrier phase and has high bandwidth efficiency,so it can be widely used in wireless communication systems.However,in the multiuser system,the non-orthogonality between the signals will cause multiple access interference,which will greatly affect the system performance and reduce the system capacity when the number of interference users is large or the system has far and near effect.This thesis focuses on the problem of multiple access interference suppression in multiuser NMM systems,and proposes several noncoherent multiuser detection algorithms which can effectively eliminate the interference.The main work of this thesis is as follows:1.Based on the pseudo-linear model of multiuser NMM system,several noncoherent multiuser detection techniques such as noncoherent decorrelation,noncoherent MMSE and sub-signal set detector are studied.An improved sub-signal set detection algorithm is proposed,which reduce the dimension of detector by using the initial solution,it can avoids the redundant interference introduced by the pseudo-linear model and greatly improves the detection performance.2.A noncoherent decision feedback detector based on QR decomposition is proposed.It uses the feedback information to cancel the interference,which can effectively eliminate the multiple access interference between users,and is suitable for multiuser systems with far and near effects.In order to further eliminate the correlation between the signals within the user,and specifically designed the linear decorrelation,the soft interference cancellation two noncoherent detection algorithms,in which the soft interference cancellation has a lower complexity by avoiding the matrix inversion.The simulation results show that the detector can greatly improve the performance of the user with weak power,and the detection performance still has good stability even if the interference user is more.3.An iterative decision feedback detector is proposed.The detector iteratively using two different structure decision feedback detection algorithms,which is based on the upper triangular matrix or based on the lower triangular matrix,it according to the maximum criteria or the mean criteria to optimize and update the user's decision estimates,can further improve the detection performance of each user in the system,the simulation results show that the detector has better performance when using the mean criterion.Aiming at the problem that the iterative detection process will increases the signal processing delay,a parallel implementation of the iterative decision feedback detector is proposed.It can not only maintain the performance of the original iterative detector,but also reduce the processing delay to the original half.