Analysis On Spectrum Sensing Techniques And Weighted Sequential Probability Ratio Test In Cognitive Radio

The radio frequency spectrum is an important and limited resource, with the rapid development of wireless communication technology, spectrum resources become more and more scarce. However, a lot of spectrum resources are idle in fact. The emergence of cognitive radio technology effectively eases the shortage of spectrum resources, and greatly improves the spectrum utilization. In cognitive radio the accurate and reliable spectrum sensing is a vital part, and this dissertation mainly analyzes spectrum sensing technology.Firstly, the dissertation introduces several single-user spectral sensing technologies and sensing data fusion algorithms in cooperative spectrum, and respectively compares their advantages and disadvantages, at the same time, gives the simulation analysis of these algorithms. Next, the collaborative spectrum sensing technology based on decision fusion is introduced, among which compares two combined rules, and gives the simulation analysis of these rules. Based on above results, the detection performance of the system is different for adopting different fusion rule. The above spectrum sensing algorithms are based on the detection of a fixed number of samples, but data fusion center also can make a final decision by not using all the node data. This dissertation describes the spectrum sensing technology of non-fixed sample-weighted sequential probability ratio test, next, for the question that malicious users sent intentionally tampering with the local spectrum sensing results to the fusion center to lead to the detection performance down, the dissertation proposes an improved weighted sequential probability ratio test algorithm based on reputation. The method effectively resists the malicious user interference and improves the detection performance of the system.