A multiobjective strategy to allocate roadside units in a vehicular network with guaranteed levels of service

Abstract

In this work, we propose the Delta-MGA, a specific multiobjective algorithm for solving the allocation of Roadside Units (RSUs) in a Vehicular Network (VANETs). We propose two multiobjective models to solve two different problems. The first one, our objectives are to find the minimum set of RSUs and to maximize the number of covered vehicles. The second one, our objectives are to find the minimum set of RSUs and to maximize the percentage of time that each vehicle remains connected. Our metric is based on Delta Network metric proposed in literature. As far as we concerned, Delta-MGA is the first multiobjective approach to present a deployment strategy for VANETs. We compare our approach with two mono-objective algorithms: (i) Delta-r; (ii) Delta-GA. Our results demonstrate that our approach gets better results when compared with Delta-r algorithm and competitive results when compared with Delta-GA algorithm. Furthermore, the main advantage of Delta-MGA algorithm is that with it is possible to find several different solutions given to the planning authorities diverse alternatives to deploy the RSUs.