Deep learning for autonomous vehicle traffic predictions in a multi-cloud vehicular network environment

Autonomous vehicles (AVs) show promise for 5G and beyond cellular networks in a variety of applications. AV utilization is rising worldwide due to the increased awareness and widespread use of artificial intelligence (AI) in numerous industries. AVs require predictive data flows to optimize data transfer and reduce latency through better utilization of transportation system capabilities, monitoring, management, and control. This research presents a novel approach utilizing a Bidirectional Long Short-Term Memory Model (BiLSTM) in deep learning (DL) to accurately forecast the traffic rate of autonomous vehicles in a Vehicular network environment that incorporates multi-cloud services. A comparison is performed between the suggested method and the traditional unidirectional LSTM for prediction accuracy as a function of batch size. According to the simulations, the suggested BiLSTM model outperforms the conventional LSTM model in terms of forecasting accuracy. Additionally, the 8-batch size model outperforms others and yields outstanding results.