Statistical adaptive heuristic model supported driver intended destination and path prediction algorithm for vehicular ad hoc network

Abstract

An emerging technology to enhance smart transportation is Vehicular Ad hoc Network (VANET) where vehicles are treated as nodes. It communicates with fixed road landed infrastructure (RSU) and other vehicles via onboard unit (OBU). Efficient and reliable information dissemination among the vehicles in sparse environment is still challenging. Also finding a shortest path from source to destination is a key factor for efficient routing. To overcome these difficulties we propose a novel framework called “Statistical Adaptive Heuristic Model Supported Driver Intended Destination and Path Prediction Algorithm (SAHDID)” which predicts the driver's indented destination and chooses a shortest time path to reach it. The routine path of each vehicle is helps us to find the most probable driver's intended destination using Hidden Markov Model. The algorithm has also enhanced further by amalgamating a heuristic algorithm which finds the shortest time route to reach the destination. The heuristic approach has perceived the lane and vehicle properties to find the candidate path. Numerical results show that the accuracy of destination prediction is about 79%. The simulation study illustrates that SAHDID outperforms well than existing VANET routing protocols in the factors of packet delivery ratio (PDR), and hop count involvement in data transmission.