WSN 6G Device-to-Device Communication

Wireless Sensor Networks (WSNs) are increasingly deployed in intelligent systems, smart cities, industrial IoT environments, and cyber security sensitive applications. However, WSNs face key challenges including energy inefficiency, network congestion, and vulnerability to routing attacks such as blackhole, grayhole, and flooding attacks. This project proposes an intelligent WSN management system integrating machine learning-driven clustering, quality-of-service (QoS) optimization, and real-time attack detection. The system uses K-Means and Agglomerative Clustering to group nodes based on efficiency parameters such as battery, RSSI, and temperature. A Random Forest-based intrusion detection module analyzes 17 network parameters to predict malicious behavior with high accuracy. QoSaware routing using Dijkstra’s algorithm ensures optimal path selection under dynamic conditions. A Flask-based backend, combined with a JavaScript visualization dashboard, enables real-time topology monitoring, cluster formation, attack alerts, and route updates. The proposed system significantly enhances network security, stability, and energy efficiency while providing an interactive platform suitable for research and real-world deployment. The system also integrates a preprocessing pipeline that enhances data quality, removes anomalies, and extracts meaningful features for accurate model predictions. Real-time simulation of sensor behavior supports dynamic updates, while the dashboard visualizes network topology, routing transitions, and attack detection through interactive animations. Performance evaluation shows improved packet delivery, reduced latency, balanced energy usage, and resilience against attacks. Together, these capabilities make the system a reliable platform for research and IoT deployments requiring secure, adaptive WSN management.