Busca e exploração de fronteiras de conectividade em redes de sensores sem fio híbridas visando integração de nós

Abstract

Wireless Sensor Networks (WSNs) integrate a wide range of applications nowadays, such as monitoring inhospitable areas, requiring the use of aerial vehicles for sensor node distribution, which can result in densely populated and sparse regions of nodes. This fact can lead to underutilization of nodes, with the emergence of nodes and sub-networks located beyond the connectivity boundary. A solution to generate greater integration is the insertion of mobile nodes into the static network, creating a hybrid network, with the aim of exploring the boundary to expand the connectivity area. However, the search process near the perimeter can result in the mobile node's disconnection due to communication loss when the node enters a network shadow area, as well as the possibility of needing to avoid obstacles, leading the node to temporarily move away from the network. These characteristics highlight the need for a control strategy for returning to the network. Due to the task's complexity, it can be segmented into subproblems or independent behaviors that interact to solve the global problem. One of the most widespread methodologies in this context is the Discrete Event System (DES), which o ers a systematic implementation with the possibility of imposing constraints on the interaction between nodes. The DES is responsible for dictating the switching rule between continuous behaviors, represented by controllers, which cannot be switched arbitrarily, as it may a ect system stability. Thus, this study aims to contribute to the creation of a navigation methodology aimed at WSNs in the process of boundary search and exploration for better utilization of distributed sensor nodes, expansion of the area of interest coverage, and allowing temporary disconnection from the network by mobile nodes. The proposed method is compared with approaches based on the Voronoi Diagram and Virtual Forces. It was veri ed that the proposed approach shows performance similar to the Voronoi approach, and both are superior to the approach based on Virtual Forces. To this end, search and exploration were explored in scenarios with low node dispersion, high dispersion, non-convex topology situations, and scenarios with obstacles. In employing the nodes, a proposed model that considers the wind e ect was used. The proposed method presents advantages in handling temporary communication loss, being restricted to navigating for a limited time under this condition and coordinated to return, the ability to avoid obstacles in the environment, and the freedom to integrate nodes in all directions and not just in the area of interest, resulting in greater node integration over longer distances traveled.