Automatic Modulation Classifiacation of MIMO Systems
No Thumbnail Available
The objective of automatic modulation classification (AMC) is to classify the modulation type of an unknown communication signal. AMC is a widely investigated area. The initial works on AMC were primarily devoted to single-input single-output (SISO) systems. There are two main approaches to AMC, namely the likelihood-based (LB) and feature-based (FB) approaches. The former treats AMC as s multiple hypotheses testing problems, where it considers L composite hypotheses H0, H1.......HL-1corresponding to L modulation types.. The multiple hypotheses testing prodedure maximizes the likelihood function of the received signal with respect to the hypotheses to decide the modulation type of the received signal The LB approach can provide optimal performances in the Bayesian sence, but it suffers from high computational complexity . The FB methods on the orher hand are not optimal in the Baynesian sense, but they require less a priori information and a low computational power. Two wodely used AMC features are the higher order mioments (HOMs) and higher order cumulants (HOCs). With the rapit development of the multiple input multiple output (MIMO) technology and its remarkable capacity gain over the traditional SISO technology, there are serious efforts to intefrate this technology into the current generation wireless systems. Designing AMC algoriths for MIMO systems is a challenging task because the mutual interference generated by MIMO spatial channels alters the statistical properties of the modulated signal. The objective of this research work is to classify the modulation fype for MIMO systems. Many studies have addressed this challenging problem. However, there remain several research issues. This thesis addresses some of these issues.
Supervisors: Prabin Kumar Bora and Ratnajit Bhattacharjee
ELECTRONICS AND ELECTRICAL ENGINEERING