IMPLEMENTATION OF CONGESTION AND COLLISION ALERT SYSTEM IN VEHICULAR AD-HOC NETWORK (VANET)

Abstract

Major challenges in urban environments include addressing the issue of traffic congestion as well as accidents at intersections. There is strong potential for Vehicular Ad Hoc Networks (VANETs) in providing real-time traffic management, collision avoidance and informative communication between infrastructure and vehicles. An effective Internet of things (IoT) hardware and VANET communication-enabled adaptive traffic management system is proposed in this paper to enhance and facilitate the flow of traffic and safety in-road travel. Using YOLO simulations, this research has proved that real-time adaptive control of traffic signals reduces traffic congestion and waiting times while improving traffic throughput by dynamically adjusting the signal timing based on actual traffic density and queue length. The above real-time system analyzes car density using surveillance cameras installed on the road using Raspberry Pi-based controllers for dynamic control of the traffic light intervals. Moreover, this system also utilizes a TCP-based VANET protocol that identifies possible collisions and sends warning messages to oncoming vehicles. Experimental results demonstrate a 35% reduction in mean waiting time at peak load and a 92% correct alert rate with an average latency of 1.2s, confirming improved flow and safety. The results confirm an effective proposition for the low-cost and scalable system in terms of better urban mobility and street safety towards smarter transport systems.