Robust defense mechanisms for protecting Wireless Sensor Networks from Denial-of-Service Threats, A comprehensive in-depth analysis

Abstract

Wireless Sensor Networks (WSNs) are inherently vulnerable due to their reliance on wireless communication and often unattended deployment. These vulnerabilities make them frequent targets of Denial-of-Service (DoS) attacks, which can significantly disrupt network operations and degrade overall functionality. This study presents a comparative analysis of the resilience of three core WSN clustering paradigms: Deterministic, Probabilistic, and Hybrid models, against a range of DoS attack scenarios. Six distinct DoS attack types, including Black Hole, various Gray Hole variants, Flooding, and Scheduling attacks are examined, along with an analysis of the impact of varying cluster head rotation periods. We provide an in-depth understanding of the impact of these attacks by examining key performance indicators like Packet Delivery Ratio (PDR), network lifetime, throughput, and energy consumption to support the development of innovative strategies for enhancing resilience. The analysis reveals that, while each clustering approach has particular advantages, there is no generic solution suitable for all scenarios. To quantify the impact of DoS attacks on clustering models, we introduced the Attack Impact Score (AIS), which measures the degradation of critical performance metrics. The Deterministic model tends to be more vulnerable to aggressive attack scenarios, whereas Probabilistic and Hybrid models show slightly greater resilience under specific conditions. However, none of the models demonstrated complete robustness when facing sophisticated attacks. Further, two-way ANOVA analysis were performed to evaluate the attack impact on key performance metrics, providing deeper insights into overall resilience of the WSN clustering architectures to various DoS attack scenarios.