ETERS: A comprehensive energy aware trust-based efficient routing scheme for adversarial WSNs

Abstract

Trust-based secure routing schemes are more effective than cryptographic routing protocols to convey energy-efficient data in WSNs since cryptographic protocols require high computation, more convergence time as well as storage space. The paper presents a well-organized trust estimation-based routing scheme (ETERS) that consists multi-trust (communication trust, energy trust, data trust) approach to alleviate several internal attacks like badmouthing, Sybil, selective forwarding, on–off,​ black hole, and gray-hole attacks for clustered WSN. The proposed multi-trust approach is used to analyze the credibility of sensitive monitored data. A novel and efficient cluster head selection algorithm (ECHSA) is employed to improve the performance of the cluster head (CH) selection process in clustered WSN. ECHSA allows the facility to elect a robust CH after a certain period to perform an equal load balance on all CHs. ETERS utilizes the Beta distribution-based trust function because recovery of trust values under attacks is faster in Beta distribution than Gaussian and Dirichlet distribution. ETERS also incorporates an irregular attenuation factor during the evaluation of communication trust to reflect the effect of various external factors such as natural calamity (earthquake), network congestion, etc., based on the trust values. However, a trust-based attack detection algorithm (TADA) assesses the reliability of SNs to detect internal attacks. TADA employs IDs, locations, and triple trust to detect internal attacks. Additionally, the proposed trust system employs an adjustable dynamic sliding length logical timing window to eradicate the limitations of existing trust models. Furthermore, a trust-based secure routing algorithm is incorporated to dynamically detect misbehavior in terms of packet forwarding for energy-efficient communication among sensor nodes. ETERS is compared with QEBSR, ATRP, ELPC, and SQEER to examine its performance. Experimental results on the MATLAB simulation platform exhibit excellent performance in terms of severity analysis, attack detection, and prevention to protect WSN, energy consumption under (i) normal (ii) attack scenario, latency (in sec.), packet delivery ratio (PDR), and throughput (in %) in the existence of spiteful nodes.