Novel Multiparameter Detection and Recovery Technique for Diverse Network Systems

Abstract

Recent research works have engendered the existence of temporary misbehaving sensor nodes in both conventional wireless sensor and modern multilayered Internet of Things (IoT) networks. This article addresses the coexistence of misbehaving things (sensors) along with properly working ones, which altogether form a diverse network. We model the performance of a diverse network using stochastic modeling and introduce a multiparameter detection technique (MDT) to detect the existence of temporary misbehaving nodes in the network. MDT offers two separate algorithms for periodic and aperiodic response networks. We also propose a simple healing mechanism to alleviate the temporary disconnections of the misbehaving nodes from the network. With a simple recovery mechanism, MDT can provide 18% extra throughput benefiting 20% less energy consumption compared to a diverse network. The proposed scheme not only alleviates the effects of network diversity, but also outperforms the other solutions like self-detection and healing.