A Deep Learning Approach for Sustainable Ad Hoc Vehicular Network

In the era of autonomous vehicles and Google cars, Vehicular Ad Hoc networks are slowly and steadily becoming a reality. V2V and V2I are two prominent and important variants of vehicular networks. RSUs are not available wherever due to any constraint like geographical terrain, so V2V is the only form available, and it should be sustainable not only for geographical constraint but also for intrusion in the network. Many AI algorithms are used for classification, but performance does not necessarily increase with an increase in data size since saturation is reached. In deep learning with an increase in data size, performance increases manifold as compared to machine learning algorithms.