QR Decomposition Multiuser Detection for Nonorthogonal Multipulse Modulation

Noncoherent decision feedback multiuser detector based on QR decomposition for multiuser nonorthogonal multipulse modulation over the additive white Gaussian noise channel is proposed. Based on QR decomposition matrix transformation, the proposed detector can effectively eliminate the multiple access interference by utilizing the decisions of stronger users. In each user, a linear decorrelating algorithm is proposed for noncoherent detection because of the multiple modulation technology. In addition, a soft interference cancellation detector is proposed to avoid the complexity of matrix inversion of decorrelating algorithm. Simulation results show that the proposed two multiuser detection algorithms can effectively eliminate the multiple access interference caused by nonorthogonal modulation, significantly improve the performance of the weak users. And the soft interference cancellation detector reduces the computational complexity by avoiding matrix inversion, while detecting performance without  loss.