Beamforming with a Simple Adaptive Partial Feedback Method in Cooperative Cognitive Radio Networks

This paper focuses on beamforming problem in a cooperative cognitive radio communication system where primary users (PUs) coexist with a secondary user (SU). An adaptive partial feedback structure for computing the channel state information is proposed. PUs and SU receivers execute their associated normalized least mean square adaptive algorithms. The output of each algorithm contains an estimate of the related channel coefficients. The receivers also reply the quantized step-size values of their algorithms via feedback channel to the transmitter, which empower it to execute these algorithms simultaneously. The transmitter iteratively employs the feedback information to estimate the channel weights and uses them for its beamforming scheme, in such a way that the SU receiver’s channel gain is maximized while the interference on PUs is maintained below a predefined level. Analysis of the proposed method shows that, with a simple modification, the required bits in feedback channel are considerably reduced. The method is applicable for time-varying systems. Besides, when the cross correlation matrix is the only available channel information between primary and secondary networks, a simple and compatible beamforming scheme is proposed. Simulation results evaluate the reasonable performance of the proposed method.