ABSTRACT
This chapter focuses on the distributed-decision fusion method by studying an optimized solution for distributed modulation classification. It explores likelihood ratio test (LRT) based Automatic modulation classification (AMC) approaches, but almost all known AMC algorithms can be extended to distributed AMC framework. The chapter reviews the LRT-based modulation classification and formulate its performance evaluation method. It explains a new method of LRT-based modulation classification, which can work well in the context of sensor networks under the constraint on the communication bandwidth. The chapter also explains new classification procedure in detail, and presents numerical results as well to show the performance enhancements by sensor networks with different structures. It discusses the application of the distributed-decision fusion rule to the problem of automatic modulation classification. An LRT-based modulation classification algorithm is pre-executed in local sensors. The sensors will not induce a potential negative contribution to the classification performance when they are located in a hostile environment.
