Clone Attack Detection Protocols in Wireless Sensor Networks: A Survey

Abstract

Wireless Sensor Network (WSN) is a collection of autonomous sensor nodes which are low cost hardware components consists of sensor nodes with constraints on battery life, memory size and computation capabilities to monitor physical (or) environmental conditions. WSN is deployed in unattended and unsecure environments, so it is vulnerable to various types of attacks. One of the physical attacks is node replication attack (or) clone attack. An adversary can easily capture one node from the network and extract information from captured node. Then reprogram it to create a clone of a captured node. Then these clones can be deployed in all network areas, they can be considered as legitimate members of the network, so it is difficult to detect a replicated node. WSN can be either static (or) mobile, in that centralized and distributed clone attack detection methods are available. In this paper we analysis various centralized and distributed protocols in the static and mobile environments. We review these protocols and compare their performance with the help of witness selection, communication and memory overhead, detection probability of replicated nodes, resilience against adversary's node compromise. General TermsClone attack, Clone attack detection approach, Static nodes, and Mobile nodes.

A Dense Neural Network Approach for Detecting Clone ID Attacks on the RPL Protocol of the IoT

“…Over time, many vulnerabilities have been reported, which have been fully or partially remediated, according to their level of risk. One of the most difficult threats to counter is the Clone ID Attack, as, according to [ 25 ], this type of threat is independent of the applications supported in WSNs, where stealthily malicious actors are able to impersonate multiple identities in the absence of default authorization mechanisms. With the rise of the information security landscape, corrective controls have been proven to be not entirely effective, as centralized and distributed methods are needed, including the use of cryptographic algorithms, hardware modifications, high-cost resources to evaluate sensor nodes positions, neighbourhood analysis and, above all, the ability to operate in restricted environments, such as IoT.…”