ENERGY DETECTION AND CYCLIC PREFIX BASED CORRELATION DETECTION FOR COOPERATIVE SPECTRUM SENSING

Abstract

Cognitive radio is a technology that provides the unlicensed users access to the bands of spectrum that is not temporarily utilized by the licensed users. These unused bands are known as spectrum holes. The process of sensing the environment to identify these spectral holes is known as spectrum sensing. Cooperation among unlicensed users for spectral sensing can be used to combat the hidden terminal problem and fading. In this dissertation work, cooperative spectrum sensing based on energy detection and cyclic prefix based correlation detection is studied for an OFDM primary system. For both these detection methods, data fusion is used at the fusion center for cooperative spectrum sensing. The receiver operating characteristics is studied for fading and shadowing channels. Data fusion is applied by maximizing modified deflection coefficient for energy detection and by Neyman-Pearson likelihood ratio test for cyclic prefix based correlation detection. The average number of statistics required for these fixed sample size tests are compared with the sequential probability ratio test