An Autonomous Fault-Awareness model adapted for upgrade performance in clusters of homogeneous wireless sensor networks

• Published in: Wireless networks 2020-10, Vol.26 (7), p.5085-5100
• Main Authors: Elsayed, Walaa M., El-Bakry, Hazem M., El-Sayed, Salah M.
• Format: Article
• Language: English
• Subjects: Adaptive filters; Communications Engineering; Computer Communication Networks; Control theory; Crashes; Electrical Engineering; Energy costs; Engineering; Fault diagnosis; Feedback loops; IT in Business; Networks; Performance evaluation; Progressions; Radio communications; Sensors; Signal monitoring; Wireless communications; Wireless networks; Wireless sensor networks
• ISSN: 1022-0038; 1572-8196
• DOI: 10.1007/s11276-020-02381-5

Wireless sensor networks (WSNs) have conquered comprehensive survey progressions in the regular control and management fields. Although WSN allows the spatial monitoring of real-world events, the mobility action depletes a huge part of a sensor’s energy cost in wireless communication. WSN sensors are often prone to various faults as frequent crashes and temporary or permanent failures. This is because it propagates them in very complex and harsh environments. So, we tend to design a Self-Adaptive based Autonomous Fault-Awareness (SAAFA) model, to limit the impact of such failures and filter them. In this paper, we incorporate the two of adaptive-filters FIR with RLS through three adaptive two-stages performed at the level of cluster head, for independent fault-correction during the propagation platform. The proposed model (SAAFA) included two stages, the first stage comprised self-detection the failure and self-aware for the lost scales, in which relied on responses of delay port and prior-knowledge of absent sensor-signals throughout monitoring, through adjusting the filter weights in the adaptive feedback loop for awarding convergent signals for the lost ones. The second stage is adaptive filtering the registered signals from the above stage for gaining pure measures and free of interferences. Compared to the state-of-the-art methods, the scheduled model attained a speed in diagnosing faults and awareness the missing readings with a rate of accuracy reached 98.8% improving the robustness of performance. Evaluation criteria revealed the progress of SAAFA in reducing the radio communication to ~ 97.47% that kept about 93.7% of battery-energy throughout the picked dataset sample. Hence, it expanded the whole network lifetime.