Multi-user detection with dynamic channel assignment

We apply multi-user detection (MUD) techniques to a TDMA system with dynamic channel assignment (DCA). With help of the DCA, the channel covariance matrix in the MUD is avoid being ill-conditioning. On the other hand, the MUD can help the DCA to improve the orthogonality between channels of co-channel users by decorrelation. Therefore, the MUD and the DCA are compensatory to each other. In addition, to reduce the complexity, only partial information of co-channel interference is used in the MUD and a succinct channel assignment rule is used in the DCA.. With different channel assignment criteria, trade-offs between network capacity and the BER performance are shown in the simulations. The proposed algorithm are shown to perform significantly better in terms of either the network capacity or the BER performance.