Network Traffic Prediction and Analysis in VANET Using Deep Learning Techniques: A Survey

Abstract

Recent advancements in information technology, intelligent public transit, and computer systems have thrown up a plethora of intelligent options for roadway security, convenience, and profitability. Among the main concerns of Intelligent Transportation Systems are the minimization of communication latency between vehicles and remote sensing units, as well as the smooth operation of traffic flow. For such issues, the Vehicular Ad hoc Network (VANET) has received interest from many research groups. Furthermore, network traffic patterns, particularly in Vehicular networks, exhibit extremely complicated behaviour due to a variety of variables such as device portability and network variability. Deep learning (DL) is being successfully used to analytics and information discovery. This study gives an in-depth examination of applications for DL in Network Traffic Prediction and Analysis (NTPA). We begin by providing basic context for our assessment. Then, we discuss the intersection of deep learning and NTPA, as well as deep learning methodologies suggested for NTPA applications. In conclusion, we explore important hurdles, unresolved concerns, and future research objectives for NTPA applications that use deep learning.