Resilient wireless sensor networks
- Publication Type:
- Issue Date:
With the increase in wireless sensor networks’ (WSN) applications as the result of enhancements in sensors’ size, battery-life and mobility, sensor nodes have become one of the most ubiquitous and relied-upon electrical appliances in recent years. In harsh and hostile environments, in the absence of centralised supervision, the effects of faults, damages and unbalanced node deployments should be taken into account as they may disturb the operation and quality of service of networks. Coverage holes (CHs) due to the correlated failures and unbalanced deployment of nodes should be considered seriously in a timely manner; otherwise, cascaded failures on the rest of the proximate sensor nodes can jeopardise networks’ integrity. Although different distributed topology control (TC) schemes have been devised to address the challenges of node failures and their dynamic behaviours, little work has been directed towards recovering CHs and/or alleviating their undesirable effects especially in Large Scale CHs (LSCH). Thus, devising CH recovery strategies for the swift detection, notification, repair and avoidance of damage events are important to increase the lifetime and resiliency of WSNs and to improve the efficacy and reliability of error-prone and energy-restricted nodes for many applications. In this research, the concepts of resiliency, fault management, network holes, CHs, TC schemes and stages of CH recovery are reviewed. By devising new TC techniques, CHs recovery strategies that partially or wholly repair LSCHs and increase the coverage of WSNs are presented such that a global pattern emerges as a result of nodes’ local interactions. In this study, we propose (1) CH detection and boundary node (B-node) selection algorithms, which B-nodes around the damaged area self-select solely based on available 1-hop information extracted from their simple geometrical and statistical features. (2) A constraint node movement algorithm based on the idea of virtual chord (v-chords) formed by B-nodes and their neighbours to partially repair CHs. By changing each B-node’s v-chord, its movement and connectivity to the rest of network can be controlled in a distributed manner. (3) Fuzzy node relocation models based on force-based movement algorithms are suitable to consider the uncertainty governed by nodes’ distributed and local interactions and the indefinite choices of movements. (4) A model of cooperative CHs recovery in which nodes move towards damaged areas in the form of disjoint spanned trees, which is inspired by nature. Based on nodes’ local interactions with their neighbours and their distances to CHs, a set of disjoint trees around the CH spans. (5) A hybrid CH recovery strategy that combines sensing power control and physical node relocation using a game theoretic approach for mobile WSNs. (6) A sink-based CH recovery via node relocation where moving nodes consider the status of sink nodes. The proposed node relocation algorithm aims to reduce the distances of moving nodes to the deployed sink nodes while repairing the CHs. The results show that proposed distributed algorithms (1)-(6) either outperform or match their counterparts within acceptable ranges. The significances of proposed algorithms are as follow: Although they are mainly designed base on the available 1-hop knowledge and local interactions of (autonomous) nodes, they result in global behaviours. They can be implemented in harsh and hostile environments in the absence of centralised operators. They are suitable for time-sensitive applications and scenarios with the security concerns that limit the amount of information exchange between nodes. The burden of decision making is spread among nodes.
Please use this identifier to cite or link to this item: