A Dynamical Relay Node placement Solution for MANETs

Abstract

Network deployment in wireless networks implies the distribution of the communication nodes to improve some key operational aspects, such as energy saving, coverage, connectivity, or simply reducing the network cost. Most node placement approaches are focused on static scenarios like WSNs, where the topology of the network does not vary over time. Nevertheless, there exist certain situations in which the network node locations can continuously change. In this case, the use of special nodes, so-called Relay Nodes (RNs), contributes to supporting, maintaining or recovering communication in the network. The present work introduces a multi-stage dynamical RN placement solution to lead the RNs to their time-varying optimized positions. The approach, named Dynamical Relay Node placement Solution (DRNS), is based on the use of Particle Swarm Optimization (PSO) algorithms and is inspired by Model Predictive Control (MPC) techniques following a bi-objective optimization procedure, where both network connectivity and throughput are jointly maximized. DRNS is validated in both simulated and real environments composed of mobile robotic nodes, the results showing its goodness and operational suitability for real MANET environments.