JoWUA

Abstract

The new paradigm of the Internet of Things (IoT), Mobile Crowd Sensing (MCS), can be used to manage traffic congestion, deliver more convenient services, and relieve the issues associated with traffic. Today, the roads with the lowest capacity and the oldest infrastructure cannot accommodate the volume of cars that flow, which results in vehicular congestion. Most traffic congestion happens during peak hours, between eight and ten in the morning, when people are going to their places of employment when students are attending educational institutions, and between four and eight in the evening when they return home. Hence this paper proposed Machine learning-based Vehicle Traffic Patterns Prediction Model (ML-VTPM) for urban transportation systems accompanied by cloud-assisted MCS architecture. Internet of Things (IoT) based sensing data devices gathered continually from many cell phones carried by drivers provide cloud-assisted MCS with the ability to regulate traffic congestion. The MCS architecture can make real-time predictions about traffic based on the information gathered from smartphones (such as speed, direction, and position). After that, the K-means algorithm makes it possible to partition the traffic into smaller groups using clustering. The weights of each cluster are then computed using the convex hull technique. The proposed ML-VTPM technology can correctly calculate the route, allowing for the fastest travel time. An offline technique based on machine learning models to forecast the mobility of vehicles in the days and weeks to come. Compared to conventional systems, the collected findings show that the suggested system offers a shorter distance and time for travel in various traffic circumstances.