On the Social Properties of Mobility Models: a Genetic Algorithm-based Approach

The performance of routing protocols in MANETs is significantly influenced by the mobility of wireless devices. Recently, many social based mobility models generate synthetic traces to capture the statistical properties detected from real traces, leaving the cause of these movement patterns untouched. As a highlighted novelty of this paper, we explore the driving force of complicated social behaviors from an evolutionary point of view and propose a genetic algorithm-based mobility model (GAMM). Using "less movement cost, more social gains" as the metric of trace's fitness to the environment, Levy Walk and inter-contact time with truncated power-law distribution emerge from generations of evolutions, accompanied by the reduction of location entropy. Two practical extension models and a simulation experiment are also presented to show the expandability and scalability of GAMM.