Research On Spectrum Sensing Algorithm Of Cognitive Radio

With the continuous development of communication technology, the problem of lack of spectrum resource is increasing seriously, cognitive radio technology is an important way to solve the problem of scarce spectrum resources, however, the spectrum sensing technology is a prerequisite for cognitive radio technology. Currently, spectrum sensing technology research is mainly doing from the two aspects, single-user sensing and cooperative sensing. The main content of this paper is on the improving algorithm based on traditional single-user sensing and centralized cooperative sensing.In the research of cognitive radio, the current research of single-user spectrum sensing mainly focused on energy detection, matched filtering detection and cyclostationary feature detection. Matched filtering method need to know a priori information of main user, so the application is relatively narrow, many experts and scholars combined the advantages of energy detection and Cyclostationary feature detection, but most of them are using a fixed double threshold to do the research, it is a lack of flexibility to the complex signals in the cognitive radio network environment and the stability is poor. To solve this problem, this paper give a joint of detection based on adaptive dynamic double threshold detection and Cyclostationary feature detection. In this improved algorithm, the cognitive users set the double threshold dynamically based on the SNR, when the signal energy of cognitive user received is not between in the double threshold, the cognitive user choose energy detection method; otherwise, the cognitive user choose cyclostationary feature detection. Finally, through experimental simulation and performance analysis, under the same false alarm probability, detection probability of the improved algorithm is more than that of the traditional fixed threshold algorithm, and the noise immunity ability is improved at the same time.In the traditional centralized cooperative spectrum sensing, the main methods of data fusion are hard fusion and soft fusion, in the soft fusion, the data of controlling channel is very excessive and the hardware implementation is difficult. In the hard fusion, mainly based on the "and" criteria, "or" criteria and "K" rank criteria, the former two fusion algorithm are inflexible and the application is more limited. Based on the traditional "K" rank algorithm, K value is fixed, when the number of cooperative users changes, it is no longer applicable, the result is either the computational complexity of the fusion center increases or the probability of detection system reduces. To solve this problem, this paper presents an optimal "K" values in the algorithm, to find the optimal value of K based on the number of cooperative users, K value appropriately increases or reduces according to the number. Finally, simulation experiments prove that under the same false alarm probability, the detection probability of this algorithm is higher than that of traditional "K" rank algorithm has, and when the SNR decreases, the detection probability of this algorithm can still keep higher.