A New Hybrid Deep Learning Algorithm for Prediction of Wide Traffic Congestion in Smart Cities

The vehicular adhoc network (VANET) is an emerging research topic in the intelligent transportation system that furnishes essential information to vehicles network. Nearly 150 thousand people are affected by road accidents must be minimized, and improving safety required VANET. prediction of traffic congestions plays a momentous role minimizing roads management for people. However, dynamic behavior degrades rendition deep learning models predicting congestion on roads. To overcome problem, this paper proposes new hybrid boosted long short-term memory ensemble (BLSTME) convolutional neural (CNN) model powerful features CNN with BLSTME negotiate vehicle predict effectively extracts from images, proposed trains strengthens weak classifiers congestion. developed using Tensor flow python libraries tested real scenario simulated SUMO OMNeT++. extensive experimentations carried out, measured performance metrics likely accuracy, precision, recall. Thus, experimental result shows 98% 96% 94% results complies clobbers other existing algorithms furnishing 10% higher than terms stability performance.