Reducing travel time in VANETs with parallel implementation of MACO (Modified ACO)

Abstract

Routing plays a major role in VANETs by helping a vehicle to reach the destination by finding an optimal path. These routing decisions are affected by the congestion on roads. Several approaches have been proposed to improve this problem of handling congestion thorough various traffic management strategies. ACO is being used in literature to provide routing in real time environment. Modified Ant Colony Optimization (MACO) algorithm is used to reduce the travel time of the journey by avoiding congested routes. This paper proposes a parallel implementation for MACO algorithm in order to further reduce the travel time due to faster computation using GPUs for the vehicles on move. Parallel implementation is done using parallel architecture on the Graphics Processing Unit (GPU) at NVIDIA GeForce 710 M using C language running CUDA (Compute Unified Device Architecture) toolkit 7.0 on Microsoft Visual Studio 2010. The obtained results for proposed parallel MACO when compared with the parallel implementation of the standard Dijkstra algorithm and that of the existing MACO algorithm on a real world North-West Delhi map with an increased number of vehicles significantly reduce the travel time.