The use of Wireless Sensor Networks (WSNs) is becoming a popular solution in dynamic environments, e.g., disaster areas, battlefield surveillance, and industrial Internet of Things. Nonetheless, node mobility, irregular topology changes, and the cyclic nature of the network’s energy consumption significantly affect network stability and data quality. Traditional energy management and failure recovery plans apply to both stationary and semi-stationary networks and therefore cannot address mobility constraints. The combination of high velocity, instability in the links, and uneven power use results in network partitioning, a low packet delivery ratio, and a shorter operational lifetime due to node failures. Moreover, existing mobility-sensitive protocols do not adequately anticipate energy scarcity and fault tolerance, leading to communication overhead and recovery delays. To address such challenges and achieve optimality in these implementations, we propose a new design, RECAST-Mob (Resilient Energy Allocation in Constrained Topology under Mobility), that leverages three suggested enhancements: predictive mobility-aware clustering, adaptive energy redistribution, and fault-tolerant multi-path routing. The Predictive Mobility Model (PMM) adopted by REACT-Mob predicts node movement and dynamically stabilises a cluster head based on residual energy, mobility factor, and link reliability, using a multi-tier energy credit system. Nodes with energy-limited offload sensing and communication to their high-energy neighbours. The tenets that are most vital to achieving resilience in their framework are lightweight beacon-based fault discovery and redundant routes that ensure data delivery continues uninterrupted when a node or link fails. There are also optional advantages of mobile sinks, such as replenishing the cluster heads when convenient and collecting information effectively, reducing communication strain. A reinforcement learning layer is also used to optimise clustering, routing, and energy scheduling to ensure that a cluster is executed optimally. REACT-Mob simulation testing has demonstrated that REACT-Mob can provide a 42% longer network lifetime, a 30% higher packet delivery rate, and a 50% faster fault recovery rate than current mobile-conscious WSN protocols. REACT-Mob is a holistic approach to designing real-time mobile WSNs that are scalable, lightweight, and sustainable, while addressing energy issues and aiming to prevent faults in the future. Its immunity to interference and its efficiency in sensing and communication make it appropriate for mission-critical applications that require standard sensing and communication.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

A Resilient Energy Allocation Framework for Fault-Tolerant Wireless Sensor Networks under Mobility Constraints

  • F. Rahman,
  • Aakansha Soy

摘要

The use of Wireless Sensor Networks (WSNs) is becoming a popular solution in dynamic environments, e.g., disaster areas, battlefield surveillance, and industrial Internet of Things. Nonetheless, node mobility, irregular topology changes, and the cyclic nature of the network’s energy consumption significantly affect network stability and data quality. Traditional energy management and failure recovery plans apply to both stationary and semi-stationary networks and therefore cannot address mobility constraints. The combination of high velocity, instability in the links, and uneven power use results in network partitioning, a low packet delivery ratio, and a shorter operational lifetime due to node failures. Moreover, existing mobility-sensitive protocols do not adequately anticipate energy scarcity and fault tolerance, leading to communication overhead and recovery delays. To address such challenges and achieve optimality in these implementations, we propose a new design, RECAST-Mob (Resilient Energy Allocation in Constrained Topology under Mobility), that leverages three suggested enhancements: predictive mobility-aware clustering, adaptive energy redistribution, and fault-tolerant multi-path routing. The Predictive Mobility Model (PMM) adopted by REACT-Mob predicts node movement and dynamically stabilises a cluster head based on residual energy, mobility factor, and link reliability, using a multi-tier energy credit system. Nodes with energy-limited offload sensing and communication to their high-energy neighbours. The tenets that are most vital to achieving resilience in their framework are lightweight beacon-based fault discovery and redundant routes that ensure data delivery continues uninterrupted when a node or link fails. There are also optional advantages of mobile sinks, such as replenishing the cluster heads when convenient and collecting information effectively, reducing communication strain. A reinforcement learning layer is also used to optimise clustering, routing, and energy scheduling to ensure that a cluster is executed optimally. REACT-Mob simulation testing has demonstrated that REACT-Mob can provide a 42% longer network lifetime, a 30% higher packet delivery rate, and a 50% faster fault recovery rate than current mobile-conscious WSN protocols. REACT-Mob is a holistic approach to designing real-time mobile WSNs that are scalable, lightweight, and sustainable, while addressing energy issues and aiming to prevent faults in the future. Its immunity to interference and its efficiency in sensing and communication make it appropriate for mission-critical applications that require standard sensing and communication.