The Research On Spetrum Sensing Technology In Cognitive Radio

Cognitive radio (CR) technique is a new paradigm of wireless communicationsystem which aims to enhance the utilization of the spectrum. This thesis has mainlyresearched spectrum sensing technology for CR Network,including single spectrumsensing and cooperative spectrum sensing. According to the research, this thesis can beconsist of three parts:In the first part of this paper，the research background and research developmentare reviewed. And then several typical spectrum sensing algorithms of single user aredescribed and compared in the aspects of computational complexity, relevance and soon. Among these algorithms the classical energy detection is researched in detail andsimulated under Additive White Gaussian Noise channel and Rayleigh channel.At the second part of this paper information fusion algorithms of cooperativesensing has been conducted. Begin from the existing Chair-Varshney algorithm and theDempster-Shafer fusion algorithm, based on the analysis of the algorithms characteristic,this paper has proposed a kind of optimized linear weighting sensing algorithm based onNayman-Pearson. This algorithm does not need the apriori information and transformvarious weighing value as a multivariable convex function under the false alarmprobability constant. Finally the simulation under the different SNR on AWGN channel,has confirmed algorithm validity.In cooperative spectrum sensing the existing linear decision fusion algorithms haveenhance the spectrum detection performance to a certain extent. However when thedetection performance of single-users drops suddenly because of the time-variablechannel or the Second User moving will also cause the overall detection performanceworsening. Based on this, This article proposed a kind of auto-adapted weightedalgorithm based on the Shapley value. This algorithm using the SNR of single-userassigned the different weight to Reducing the influence. The simulation indicated: Thisalgorithm has enhancement performance, compared with the traditional linear decisionfusion algorithms And it has the very good actual utilization prospect.