A Neural Network Blind Multi-user Detection Algorithm

Abstract

This paper proposed a blind multi-user detection algorithm based on neural network. The constant modulus cost function was constructed and cost function with restricted condition was optimized by augmented Lagrange method. Iterative formula of neural network weight and parameters were deduced by Least Mean Square. Blind multi-user detection can be realized by iterative functions of neural network. Simulation indicates the new algorithm improves the ability of anti-MAI., reduces error rate, and speeds up convergence rate.