A Novel Approach for Real Time Traffic Prediction Using Deep Learning in VANET

VANETs, or vehicular ad-hoc networks, are the upcoming technologies in vehicles for better and safer transportation using wireless communications. These intelligent vehicles can communicate with Roadside Units (V2R) as well as with other vehicles (V2V). With the increase in growth of the population, especially in metropolitan cities, one of the major concerns faced is increase in traffic. VANETs provide efficient traffic management by communicating between vehicles. Rapid changes in current technologies provided us with many new solutions for existing problems and one such technology is Deep Learning. We can deploy Deep Learning models to identify the pattern of vehicle travelling at a specific region which helps us in scheduling the vehicles thereby reducing traffic. Time series forecasting models such as Recurrent Neural Networks (RNN), Bidirectional Long Short-Term Memory (Bi-LSTM), Gated Recurrent Units (GRU) etc., uses the past data to predict the present and future outcomes. Different Neural Networks as mentioned above are used in the present work and a Hybrid model is developed combining Convolutional Neural Network (CNN) along with LSTM. These architectures are compared and Hybrid model developed outperformed other models with the accuracy of 99.96%.