(Research work)
Papers & experiments
Peer-reviewed work across post-quantum cryptography, security and AI. Scroll the shelf — open any card for the summary and DOI.
CMSS-PPR: A Post-Quantum Multivariate Signature Scheme
Scopus-indexed journal (under review)
A new post-quantum signature scheme built on Non-Affine Ring-Permutation Layers (NARPL) that raise the system's regularity degree — shutting down MinRank and Gröbner-basis attacks. Its security reduces to MQ hardness, coming out algebraically stronger than Rainbow, UOV and HFE.
Advancements in Isogeny-Based Cryptography: A Mathematical Approach to Post-Quantum Security
IEEE SCEECS 2025 · Bhopal
A deep dive into isogeny-based cryptography as a post-quantum candidate. We pin down where SIDH and SIKE fall short on security and speed, then introduce a new isogeny pathfinding algorithm — using graph theory and probabilistic models — that cuts computational complexity while improving both efficiency and security.
AI-Enabled Posthumous Identity Protection Using Advanced Cryptographic and Governance Mechanisms
IEEE SCEECS 2026 · Bhopal
What happens to your digital identity after you're gone? The Digital Afterlife Protection Framework (DAPF) secures a deceased person's accounts and data using blockchain records, asymmetric signatures, multi-factor identity checks and anomaly-based fraud detection — with an ethical, user-driven choice over what happens to each asset.
Agentic AI Framework for Autonomous and Self-Managing Cloud Services
IEEE SCEECS 2026 · Bhopal
Rule-based Kubernetes autoscaling struggles when workloads shift unpredictably. This framework uses multi-agent reinforcement learning so autonomous agents orchestrate cloud-native apps together — beating traditional orchestration on resource usage, latency and resilience in simulation.
RF-Driven Security Analytics for Intrusion Detection in Critical Defense Systems
IEEE ISCS 2025 · Gurugram
Traditional intrusion detection trusts a network that may already be compromised. This system watches radio-frequency side-channel emissions instead — passive, air-gapped software-defined radios pick up spectrum anomalies to spot intrusions in critical defence systems without ever touching the host.
Dynamic Browser Threat Detection Leveraging Multi-Engine Threat Analytics
IEEE SCEECS 2026 · Bhopal
Antivirus tools react after the fact; this catches malware before it lands. A lightweight Chrome extension scans every downloadable link against multi-engine threat analytics (VirusTotal) plus a whitelist, flagging malicious files before the download even starts.
A Cloud-Native Management System Using Serverless Architecture
IEEE SCEECS 2026 · Bhopal
A global, real-time campus event platform built serverless on Next.js and Firebase. Real-time discovery and registration, QR-code ticketing, AI-based recommendations, cross-college collaboration and payment gateways — fast, responsive and GDPR-compliant by design.
A Resilient Peer-to-Peer Validation Framework for Reliable Digital Coupon Utilization
IEEE SCEECS 2026 · Bhopal
Most digital coupons expire unused. This peer-to-peer platform lets people exchange coupons anonymously and safely, backed by a trust-and-reputation mechanism, image-verified uploads and abuse prevention — built on the MERN stack for real-time, large-scale sharing.
The Future of Digital Payments: Addressing Security Challenges Through Cryptography
Springer · Book Chapter
Digital-payment rails are fast and convenient — and a much bigger attack surface, now including the looming quantum break. We propose a multi-layer crypto scaffold combining hybrid and post-quantum encryption, threshold (secret-sharing) signatures and live fraud detection — plus a mobile/NFC-friendly isogeny signature — delivering forward secrecy and quantum resistance at real transaction speed.
Hybrid Spectral-Spatial Graph Neural Networks: Bridging Spectral Theory and Scalability
Springer · Book Chapter
A Graph Neural Network that unifies spectral and spatial methods in one architecture — Laplacian eigenstructure for global graph geometry, localized message-passing for fine relational detail. It eases the usual expressiveness-vs-scalability trade-off, beating strong baselines on Cora, PubMed and ZINC with lower compute.