Suggestion of routes for vehicles in vehicular networks using the multicommodity flow model

This work models the traffic congestion minimization as the multicommodity flow problem, which minimizes the amount of vehicles that travels in the same streets. Based on two constructive heuristics proposed in this study (Random Routes and Sorted Routes), we have developed and evaluated four new heuristics, being two local search and two metaheuristics of reactive GRASP. The results have shown that heuristics based on Random Routes have better performance, reducing the travel time in up to 56%, the CO2 emissions in up to 18% and fuel consumed in up to 18%, in average.