ANALYSIS OF DECISION THEORETIC MODULATION CLASSIFICATION METHODS FOR DIGITAL COMMUNICATION SIGNALS

ABSTRACT computationally intensive but current microprocessors have made the ALRT practicable. In general, the Automated modulation classification is a performance of ALRT AMC is very sensitive to fundamental requirement for electronic support modulation parameters such as symbol timing, baud rate, measures. Existing automated classifiers use a variety of carrier frequency, carrier phase, pulse shape, and noise different modulation recognition techniques. This paper power. The ALRT is also affected by channel fading and reviews the category of decision-theoretic approaches the type of noise that is present. and discusses the relationships between decision-theoretic methods and other statistical modulation The GLRT (Panagiotou et al., 2000) treats the classification methods, candidates as unknown deterministic values and the maximum likelihood test is applied as if the true values 1. INTRODUCTION were known. The HLRT (Panagiotou et al., 2000) is a hybrid approach that treats some of the candidate The goal of modulation classification is to take an parat asprano variabs with ow pdf and unknown signal and identify the modulation type with a sometofsthe candia para mes as un