DefinePK

Abstract

Increased demand for technology-driven and automated infrastructure that can address critical issues like passenger safety and traffic congestion has led to an exciting research and application area - Vehicular Adhoc Network (VANET). VANET enables vehicles to talk among them and also with fixed roadside infrastructure to support a myriad of potential life changing applications. The excitement surrounding VANET is not due to their application support or potential benefits but also because of the challenges like scarce spectrum, varied QoS requirements, poor connectivity, security issues etc., Cognitive Radio (CR), a technology that ensures efficient spectrum usage can be employed in VANET to address spectrum scarcity issue. Though several spectrum sensing algorithms have already been proposed, there is a need for an effective algorithm that has a significant impact on various sensing parameters like accuracy, delay and efficiency. Our focus in this paper is to provide a machine learning based sensing algorithm for CR VANETs implemented at physical layer that maximizes the spectral efficiency.