Sequence design for MPG QS-CDMA systems based on heuristic combinatorial optimization

This paper deals with a quasi‐synchronous code division multiple access (QS‐CDMA) system with a multiple processing gain (MPG) variable data rate scheme in the uplink direction and subject to multipath fading channels. An analytic expression is obtained for the signal‐to‐noise plus interference ratio (SNIR) at the Rake receiver output (finger). In order to maximize the SNIR, three combinatorial optimization methods applied to sequences selection are compared. Then, sequence selection methodology for this class of system is evaluated, comparing the local search optimization results with unitary Hamming distance (1‐opt LS), evolutionary programming with cloning (EP‐C), and simulated annealing (SA) ones. The contribution of this work consists of considering the self‐interference (SI) effect, besides the multiple access interference (MAI), in the description of the proposed objective function for sequence selection and a compared analysis of three heuristic algorithms maximizing that SNIR. Also, numeric results have demonstrated the effectiveness of the proposed method for the QS‐CDMA sequences selection. Copyright © 2010 John Wiley & Sons, Ltd. A heuristic methodology for sequences selection suitable for multirate MPG QS‐CDMA systems under multipath fading channels is proposed in this paper. Based on the specific SNIR target for each class of service, the suggested objective function was shown appropriate for this problem. Under distinct QoS (satisfied if fo(A) < 1), the best convergence time among the three combinatorial optimization algorithms was obtained with the EP‐C. However, the smallest value of the objective function was achieved by the SA.