Fault Detection and Performance Improvement in Wireless Sensor Networks Using the Fault Detection Node Scenario (FDNS) Approach

DOI

https://www.doi.org/10.59256/indjcst.20260503004

Published In

Volume 5, Issue 3 (2026)

Published Online

September-December 2026

Pages

18-25

Article Type

Original Article

Check for updates

Wireless Sensor Networks (WSNs) are heavily deployed in dynamic, unpredictable, and harsh environments for real-time monitoring and control. In these mobile scenarios, low-power sensor nodes are highly susceptible to physical transceiver failures and logical anomalies, such as localized packet dropping and geographic coordinate misreporting (spoofing). Traditional centralized or deep learning-based diagnostic frameworks impose unsustainable processing, memory, and energy footprints on resource-constrained microcontrollers, leading to high packet latencies and premature network exhaustion. To resolve these challenges, this paper presents Fault Detection Node Scenario (FDNS) scheme. The proposed model shifts the burden of fault diagnostics from a centralized Base Station (BS) to local, one-hop neighboring sensor nodes. Neighbors monitor transmission rates over sliding temporal windows, isolate anomalous nodes locally, and bypass them by dynamically recalculating multi-hop routing paths. The baseline feasibility and mobility tolerance of the proposed scheme are rigorously evaluated within Network Simulator 2 (NS-2.35) across a 50-node mobile topology, with velocities systematically scaled from 2 m/s to 10 m/s. Under controlled fault-injection conditions, the unmitigated faulty scenario experiences a severe Packet Delivery Ratio (PDR) collapse to 72.87% and a sharp surge in Average End-to-End Delay to 180.95 ms. The proposed localized Fault Detection Node Scenario (FDNS) successfully intercepts this degradation. Averaged across the evaluated mobility range, FDNS restores the PDR to 81.47% (representing an 11.80% relative improvement against the faulty state), reduces the Packet Loss Ratio (PLR) by 31.70%, cuts routing control packet overhead by 11.12%, and recovers active data throughput by 14.79% while preventing the latency escalations that cripple standard routing protocols under failure.