Multiuser detection based on the immune strategy RBF network

The radial-basis function (RBF) neural network is a model of locally approaching network that is now applied widely. The learning algorithm of this model is usually paid much attention because it is a main approach to increase the working performance of RBF network. On the other hand, the evolutionary algorithm based on the immune strategy is a globally optimal algorithm that possesses batter pertinence, stronger robust and higher speed. Therefore, the combination of this model and the algorithm can generate a novel and powerful RBF network based on immunity and evolutionary. It is illustrated from the application of this novel model in CDMA that the learning algorithm can conduce to decrease the bit-error rate and increase the converging speed, and then improve the realtime capability and practicability.