Fuzzy System based Weights for Energy Detection in Cognitive Vehicular Networks

VANET is an ad-hoc network in which vehicles are connected over a wireless medium which is allocated 75MHz of bandwidth for vehicular communications. The present radio channels designated for VANET applications are a result of advancements made in wireless applications for vehicles. There are seven channels in the spectrum each with 10MHz bandwidth which are prone to congestion when numerous vehicles are aiming for the common medium. Cognitive Radio technology has been a promising solution which can improve the performance of vehicular communications. In cognitive radio network, a transceiver can discern between channels that are actively being used and those that are not. Unlicensed users can utilize the channels when there are holes in the licensed spectrum by performing spectrum sensing. Spectrum sensing in vehicular networks has many challenges, including the high mobility of cars, the existence of obstacles, shadowing effect, and the dynamic nature of the transmission medium. As a solution to these problems, a weight-based energy detection method has been proposed that makes use of a fuzzy inference system which is used for calculating weights in decision-making circumstances when the input data is uncertain or imprecise. The SNR, vehicle’s distance from PU, time, and vehicle’s speed are the parameters considered in the fuzzy system. These weights are used with the local choice made by vehicles to define the channel state more precisely. Each parameter is defined in the fuzzy system using membership functions which generates output weight value based on the reliable conditions. The simulation results produced using Matlab demonstrated improved detection performance over local spectrum sensing.