| Mobile ad hoc networks (MANETs) are increasingly used to acquire information in harsh and inaccessible two-dimensional (2 D) and three-dimensional (3D) environments. Self-deployment of autonomous mobile nodes is a critical task to efficiently control the topology of MANETs. Robust topology control algorithms must be developed to guide the mobile nodes to achieve a uniform distribution without a central controller in the absence of a priori knowledge of the field.;In this dissertation, a new self-positioning technique, node spreading voronoi-based genetic algorithm (NSVGA), for autonomous nodes in a mobile ad hoc network to spread over unknown two-dimensional deployment terrains is presented. NSVGA relies on a genetic algorithm that utilizes only the local neighbor information and the Voronoi tessellation of nodes sensing area as a fitness function. We show that NSVGA can keep the network connected and become stable in harsh environments.;We also introduce topology control algorithms, called 3D-GA and 3D-PSO, by extending existing two-dimensional algorithms for three-dimensional space, specifically for underwater environments. Each unmanned underwater vehicle (UUV), guided by our algorithms, is capable of adjusting its speed and movement direction to achieve a better location toward obtaining uniform distribution. We formally analyze the behavior of our algorithms with respect to spreading, uniformity, and connectivity metrics. Since the imprecise and limited neighbourhood knowledge could potentially disrupt convergence towards a uniform and stable spatial coverage, we test our algorithms in hostile (i.e., vehicles may experience hostility and become disabled) and noisy environments (i.e., next movement calculations contains inaccurate location information from the local neighbors). Our simulation experiments show that our algorithms are effective tools for providing a robust solution for volumetric spatial control of UUVs in MANETs.;We also extend our 3D-PSO algorithm to provide a user-defined level of protection density and fault tolerant connectivity by integrating shield density and Yao graph into our fitness calculation. By modifying the fitness function, our 3D-PSO algorithm can guide the UUVs to achieve different configurations for various applications such as protecting harbor entrances, ships and submarines. |
