Collaborative Detection For Cognitive Radio Spectrum Based On Neural Network

With the rapid development of wireless communication, more and more people pay attention to spectrum resource. In order to solve the shortage of the spectrum resource, researcher point out the technology named as cognitive radio. The technology does no any interference to the normal communication. However it can detect the spectrum which is allocated to the authorized users. The unauthorized users utilize the void spectrum to communicate when it detects the spectrum which no users use. We use the cognitive radio technology to detect the void spectrum to provide more access to opportunity for the unauthorized users. So the secondary users have no any cost. This method is more effective than the tradition communication. Due to the effect of buildings obstruction, the distance between authorized users and unauthorized users, the detection result of single node may not be accurate. So we have joint spectrum sensing method to consider, use the fusion rule to obtain result. Although it can improve the detection performance, the center station receives information from sensing nodes, which cause network load and make the network bandwidth congestion.The article applies neural network which has the characteristics of artificial intelligence and self-organizing learning in joint spectrum detection. Through neural network training, center base station obtains reliable information to participate in joint detection. In order to reduce number of sensing nodes to cooperate, lower the network load. The method reduces the congestion of information. It can improve the utilization rate of spectrum.