Study on Spectrum Sensing Techniques

Abstract

Cognitive radio is a recent emerging technology for effective use of spectrum. To provide newer opportunities in advanced wireless communication system. Spectrum sensing and security are the key issues of Cognitive Radio (CR). The main aim of Cognitive Radio (CR) is to detect unused or idle frequencies of cellular spectrum, not used by Primary User (PU’s). Practically PU’s frequencies can be idle or busy. The goal is to investigate spectrum sensing predictive methods. These methods find idle or busy states of frequencies at discrete interval of time periods. The decision is made to assign these free frequencies to Secondary Users (SU’s). For this we use Machine Learning (ML) techniques to improve its performance, usability and efficiency. A machine learning system includes pre-processing, feature extraction and decision making classifier. This paper lists recent research activities and co-operative spectrum sensing. Identification of CR terminals is done by predictive spectrum sensing methods. Machine learning algorithm are used for classification.