Reliability-Aware Clustering for IoT Device Layers based on Cold Standby Redundancy
Abstract
Smart environments, including those in healthcare, transportation, industry, as well as urban infrastructures, have deployed vast numbers of linked devices as a result of the fast expansion of the Internet of Things (IoT). The network is very susceptible to failures, energy depletion, and instability since IoT devices still have limited resources, even if their relevance is increasing. While LEACH and HEED are two examples of traditional clustering protocols, they place an emphasis on energy efficiency at the expense of dependability, which becomes more important in situations when cluster heads (CHs) fail often. This has led to the emergence of fault tolerance in IoT clustering as a major area of focus for researchers. In order to improve system dependability while maintaining energy economy, this research suggests a reliability-aware clustering model that incorporates cold standby redundancy at the cluster head level. By allowing a dormant backup node to take over following a CH failure, cold standby redundancy achieves a balance between fault recovery and energy conservation, in contrast to hot standby redundancy, which gives instantaneous recovery but uses significant energy. Three scenarios no redundancy, cold standby redundancy, & hot standby redundancy, are used to mathematically develop and test the proposed model using MATLAB-based simulations. Important indicators for performance assessment include scalability, latency, energy consumption, reliability improvement ratio, and fault recovery time. Compared to baseline clustering, cold standby significantly increases reliability while keeping minimal energy overhead, according to the simulation findings. Hot standby speeds recovery, but it's not practical in resource-constrained IoT settings due to its high energy cost. According to the scalability study, the suggested model is suitable for large-scale deployments since its overhead grows linearly. All things considered, the results show that cold standby redundancy is a viable and scalable option for the future generation of IoT networks since it provides a good compromise between dependability, energy economy, and recovery delay.