Performance metrics for self-positioning autonomous MANET nodes

Janusz Kusyk, Stephen Gundry, CemSafak Sahin, M mitUyar
2013 Journal of Cyber Security and Mobility  
We present quantitative techniques to assess the performance of mobile ad hoc network (MANET) nodes with respect to uniform distribution, the total terrain covered by the communication areas of all nodes, and distance traveled by each node before a desired network topology is reached. Our uniformity metrics exploit information from a Voronoi tessellation generated by nodes in a deployment territory. Since movement is one of the most power consuming tasks that mobile nodes execute, the average
more » ... cute, the average distance traveled by each node (ADT) before the network reaches its final distribution provides an important performance assessment tool for power-aware MANETs. Another performance metric, network area coverage (NAC) achieved by all nodes, can demonstrate how efficient the MANET nodes are in maximizing the area of operation. Using these metrics, we evaluate our node-spreading bio-inspired game (BioGame), that combines our force-based genetic algorithm (FGA) and game theory to guide autonomous mobile nodes in making movement decisions. Our simulation experiments demonstrate that these performance evaluation metrics are good indicators for assessing the efficiency of node distribution methods. Mobile ad hoc networks (MANETs) are useful for many commercial and military applications where network coverage is needed over a terrain without an established infrastructure. Autonomous topology control algorithms aim to provide a method to deploy mobile assets without a centralized controller such that MANETs are scalable and robust to node failures. In this context, it is advantageous when the MANET topology has reduced sensing overshadows, limited blind spots, enhanced spectrum utilization, and simplified routing procedures while lowering power consumed by each node. Achieving these objectives necessitate that autonomous nodes in MANETs (a) place themselves over an unknown geographical terrain in order to obtain a uniform network distribution, (b) reduce the total distance traveled before overall network objectives are reached, and (c) preserve network connectivity while attaining positions that ensure a high coverage of the area by all nodes. In this article, we present quantitative methods to evaluate performance of node selfpositioning techniques with respect to uniformity of distribution of nodes, average distance traveled (ADT) by each autonomous vehicle, and the total area coverage (NAC) obtained by all nodes. The uniform distribution of mobile nodes is often a desired network topology that helps to prolong network lifespan by ensuring that nodes deplete their energy resources evenly. When MANET nodes are uniformly distributed, they are able to equally share sensing and communication tasks, hence the likelihood that a single node ceases to function much earlier than expected due to power exhaustion is reduced. In order to gauge the performance of a MANET with respect to its uniform distribution, we define metrics based on Voronoi regions generated by the nodes. Since a node's movement is typically the most energy-consuming operation performed by autonomous vehicles, reducing the distance that nodes travel is an important objective. ADT provides a realistic metric for evaluating self-spreading algorithms where preserving scarce energy resources is imperative. Gathering information about operational environments to provide mission-critical data is often the main motivation for deployment of MANETs. In order to adequately utilize the communication coverage of deployed nodes,
doi:10.13052/jcsm2245-1439.223 fatcat:o32wuml7nnb4rdo3d7xtzytda4