An improved genetic algorithm based on eugenic population for multiuser detection in DS-CDMA systems

Although the embedded genetic algorithm and multistage detector (E-GA/MSD) can improve the performance of the conventional GA in multiuser detection, the computational load of MSD is considerable when the number of users is large. To overcome this problem, an improved GA with eugenic population is developed. Based on a new strategy of survival selection, the eugenic population is more confidential than the elitist method in the evolution of better chromosomes for the next generation. Thus, the proposed technique can achieve almost the same performance as the E-GA/MSD, but without incurring additional computations of MSD. Simulation results demonstrate the improved performance of the proposed GA.