Rapid Adaptive Beamforming Algorithm for Multi-user Cognitive Radio Systems

The paper proposes a rapid adaptive beam forming algorithm based on a modified constant modulus algorithms for multi-user cognitive radio systems. The algorithm combines the search direction based on Quasi-Newton algorithm and subspace projection method, this algorithm has quadratic convergence rate. The weight vector at each iteration locating at the signal subspace further improves the convergence speed and accuracy, which meets real time of the cognitive users, and improves the SINR performance. Simulation results show that compareed to beam forming based on the other constant modulus algorithms, the proposed algorithm converges faster, and has higher signal to interference ratio.