Software-defined decision-feedback multi-user detection in frequency domain for single-carrier and multi-carrier CDMA systems - a sequential quadratic programming approach

This paper shows how a unified model can he used to describe the signal received in frequency domain over a multiple-antenna, frequency-selective, multipath fading channel for multicarrier CDMA and single carrier CDMA systems with cyclic prefix. This realization validates the abstraction of these CDMA problems into a mathematical model which can be used for practical problems such as channel estimation and data detection. Moreover, the optimal multiuser detection problem based on the maximal-likelihood criteria for this unified model is addressed as a combinatorial optimization problem. A sequential quadratic programming approach is proposed to approximate this problem, which corresponds to a near-optimal software-defined decision-feedback multiuser receiver for both single-carrier and multicarrier CDMA systems. An efficient gradient projection algorithm is applied for data detection in the proposed receiver. The simulation results show that this scheme can reach near single-user performance even for fully-loaded systems and outperform conventional soft-interference cancellation methods.