A fair trust-based malicious node detection and isolation scheme for WSNs

Abstract

Wireless Sensor Networks (WSNs) are vulnerable to a large number of security threats where sensor nodes may be compromised by malicious nodes that could inject the network with falsified information or bring the network down. Traditional security mechanisms are not suitable for WSNs as these mechanisms require extensive computation, communication and memory, which contradict the nature of a WSN. Trust and reputation based schemes are more effective in providing additional means of security for decision making in WSNs. The main objective of such schemes is to detect and isolate the maliciously behaving nodes before they can do any harm to the network. This paper investigates the performances of a fair trust-based malicious node detection and isolation (FTMNDI) scheme. This scheme uses a modified version of the neighbor-weight trust determination (NWTD) algorithm to periodically update the trust of the nodes based on their reputation. Then, a node within the network is isolated if its trust becomes less than a pre-set minimum acceptable trust value. The performance of the FTMNDI scheme is evaluated through simulations using the network simulator MANSim. The simulation results demonstrate the effectiveness of the FTMNDI scheme in detecting and isolating malicious nodes in a fairly and timely manner.