INVESTIGATION OF SPATIAL CONFIGURATION OF SENSOR NODE PLACEMENT IN A TWO-DIMENSIONAL ENVIRONMENT

Abstract

This paper focuses on the investigation of optimal spatial deployment strategies for sensor nodes in wireless sensor networks. The proposed solution aims to achieve a balanced distribution of sensing elements across a two-dimensional environment while simultaneously reducing redundant overlaps between their coverage areas. Such an approach is essential for enhancing coverage effectiveness and ensuring efficient utilisation of network resources. The node deployment task is modelled as a complex optimisation problem in which the search domain consists of all feasible coordinate configurations of sensor nodes, considering both existing infrastructure and newly introduced elements. A key contribution of the proposed methodology lies in the application of a dedicated fitness evaluation mechanism designed to penalise local concentration of nodes within the deployment area. This enables the prevention of clustering effects and supports the formation of a more evenly distributed network topology without causing substantial disruption to the original network structure. Through extensive simulation experiments, the most favourable operating conditions of the genetic algorithm were determined. The highest deployment performance was achieved using a population of 800 candidate solutions evolved over 200 generations, with crossover and mutation probabilities selected within the intervals {0.5; 0.8} and {0.05; 0.2}, respectively. Additionally, a sensing radius of 30 m and a minimum permissible separation distance of 20 m between neighbouring nodes were found to be optimal. Examination of the fitness function behaviour across successive generations revealed that the optimisation process reaches a steady state after approximately 130 iterations, indicating reliable convergence towards high-quality solutions. The simulation outcomes demonstrate that the proposed genetic-based strategy successfully places all newly introduced sensor nodes without generating overlapping coverage regions. Comparative evaluation against uniform, greedy, and random deployment techniques highlights the clear superiority of the developed approach. Overall, the results confirm that the proposed genetic algorithm constitutes an efficient and robust solution for addressing the problem of spatial sensor node placement in wireless sensor networks characterised by partially established topologies.