SVM and Decision Stumps Based Hybrid AdaBoost Classification Algorithm for Cognitive Radios

Abstract

In this paper machine learning techniques based cooperative spectrum sensing (CSS) algorithms are investigated for cognitive radio networks (CRN). A novel support vector machine (SVM) and decision stumps based AdaBoost classification algorithm is proposed for pattern classification of the primary user’s behavior in the network. Conventionally, Ad-aBoost algorithm combines multiple sub-classifiers and produces a strong weight based on their own weights in classification. Taking into account the fact that SVM and decision stump serve as relatively strong and week classifiers respectively, the proposed algorithm employs SVM as the first-stage classifier and decision stump as the second-stage classifiers to eventually determine the class that the spectrum energy vector belongs to. It is verified in simulations that the proposed algorithm is capable of achieving higher detection probability than the conventional machine learning algorithms.