A Cross Validated Clustering Technique to Prevent Road Accidents in VANET

Vehicular Ad-hoc Network is one of the fastest growing network of this modern era and world. It has a broad variety of scope. Due to a high speed moving nodes, it becomes a little difficult to keep information about vehicles at the right time. Due to this lag, lot of data packets are dumped and the overall success rate of delivering packets becomes low. This paper emphasizes on a novel clustering concept utilizing Cluster Head mechanism and Artificial Neural Network (ANN) to enhance the data transfer mechanism. The performance of proposed algorithm is evaluated using Mean Square Error and Packet Delivery ratio. A new model of cross-validation is also considered in this scenario.