Research Assistant

Overview

As a Research Assistant working under the guidance of Prof. Sudeep Tanwar and Prof. Rajesh Gupta, my research focused on creating robust, secure, and interpretable machine learning systems for Internet of Things (IoT) and connected vehicle environments. The primary objective of our research group was to design intelligent models capable of detecting cyber threats while providing explainable decision paths to security auditors.

Key Projects & Technical Details

1. Vehicular Network Intrusion Detection (V2X Security)

• Designed and evaluated deep learning models to detect False Data Injection (FDI) and sensor spoofing attacks within Connected and Autonomous Vehicle (CAV) networks.
• Simulating realistic communication scenarios to benchmark network vulnerability against coordinated adversarial attacks.
• Implemented anomalous traffic classification pipelines capable of flagging network threats in real time.

2. Decision Auditing via Explainable AI (XAI)

• Applied SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-agnostic Explanations) to unravel the “black-box” decision process of high-performance classifiers.
• Built auditing protocols that analyze feature attribution shifts when models detect adversarial inputs, establishing baseline verification criteria for trustworthy systems.
• Refined visual explanation models to map complex sensor input importances for human auditors, helping bridge the gap between network operators and raw telemetry predictions.

Peer-Reviewed Publications

During this tenure, I co-authored 5 peer-reviewed research papers published in reputable IEEE and Springer venues. These publications focused on:

1. Quantum-secured vehicle-to-everything (V2X) communications (Q-ShielD).
2. Deep learning and explainability audits for connected vehicle sensor streams.
3. Lightweight, approximate machine unlearning algorithms for edge networks.
4. Quantum-assisted edge intelligence for real-time telemetry anomaly detection.
5. Explainable AI and quantum security frameworks forsmart home networks.