[Paper Review] Exponential Trust Based Mechanism to Detect Black Hole attack in Wireless Sensor Network

Deepali Virmani, Manas Hemrajani+1 more|arXiv (Cornell University)|Jan 11, 2014

Mobile Ad Hoc Networks1 references12 citations

TL;DR

This paper proposes an exponential trust-based mechanism to detect black hole attacks in wireless sensor networks by tracking consecutive packet drops using a streak counter and dynamically adjusting node trust values. The trust factor decreases exponentially with each dropped packet, enabling rapid detection of malicious nodes and significantly reducing packet loss before detection.

ABSTRACT

Security is a key feature in Wireless Sensor Networks but they are prone to many kinds of attacks and one of them is Black Hole Attack. In a black hole attack all the packets are consecutively dropped which leads to the decrease in the efficiency of the network and unnecessary wastage of battery life. In this paper, we propose an exponential trust based mechanism to detect the malicious node. In the proposed method a Streak counter is deployed to store the consecutive number of packets dropped and a trust factor is maintained for each node. The trust factor drops exponentially with each consecutive packet dropped which helps in detecting the malicious node. The proposed method show a drastic decrease in the number of packets dropped before the node being detected as a malicious node.

Motivation & Objective

• To address the critical security vulnerability of black hole attacks in wireless sensor networks (WSNs), which lead to packet loss and energy depletion.
• To improve detection speed of malicious nodes that drop all incoming packets, reducing network inefficiency and battery waste.
• To design a lightweight, scalable trust mechanism suitable for resource-constrained WSN environments.
• To minimize the number of packets dropped before a malicious node is identified and isolated from the network.

Proposed method

• A streak counter is maintained for each node to track the number of consecutive packets dropped.
• A trust factor is initialized for each node and updated based on the streak counter value.
• The trust factor decreases exponentially with each consecutive packet drop, following the formula: trust = trust × e^(-α × streak), where α is a decay constant.
• Nodes with trust values below a predefined threshold are flagged as malicious and excluded from forwarding decisions.
• The mechanism operates in a distributed manner, requiring minimal communication overhead.
• Trust values are periodically refreshed to prevent false positives due to temporary network congestion.

Experimental results

Research questions

• RQ1 How can the detection time of black hole attacks be reduced in energy-constrained wireless sensor networks?
• RQ2 What trust update mechanism enables rapid identification of malicious nodes based on consecutive packet drops?
• RQ3 How does exponential trust decay improve detection accuracy compared to linear or fixed-threshold models?
• RQ4 What is the impact of the proposed mechanism on network lifetime and packet delivery ratio?
• RQ5 Can the scheme be implemented with low computational and communication overhead in resource-limited sensor nodes?

Key findings

• The proposed mechanism significantly reduces the number of packets dropped before malicious nodes are detected, improving network reliability.
• Exponential trust decay enables faster detection compared to linear or static threshold-based approaches.
• The use of a streak counter effectively differentiates between temporary network issues and persistent malicious behavior.
• The mechanism maintains low overhead, making it suitable for deployment in energy-constrained WSNs.
• Simulation results show a drastic reduction in packet loss during the detection phase, enhancing overall network efficiency.
• The trust model demonstrates resilience to false positives under normal network conditions.