Performance Analysis of a 5G Wireless Network for Vehicular Communication using Machine Learning

The recent progress in wireless applications for vehicles is another significant factor in the scarcity of spectrum. The cognitive radio model is a tool that enables unlicensed cognitive users (CUs) to make use of vacant, idle bands. The fundamental element of cognitive radio networks is the quick and accurate identification of the major legacy user. However, low SNR issues brought on by shadow fading and buried terminals fundamentally limit sensing capability and have practical implications for the architecture of cognitive vehicular networks. To characterize different channel properties, especially multipath fading and also shadowing, extensive modeling is being done. For various vehicle to vehicle (V-2-V) and vehicle to infrastructure (V-2-I) communications, spectrum sensing is a practical option. This spectrum sensing is based on energy detection (ED). In order to accommodate small and large scale fading, this work investigates the spectrum sensing performance utilizing ED across Rayleigh and Gamma-shadowed Nakagami fading channel. The findings demonstrate how significantly fading severity and shadowing spread affect detection ability. We provide the pertinent simulation results using machine learning technique to back up our analytical findings.