Nsoh Research Artifacts
This gallery curates the security artifacts produced through Design Science Research Methodology (DSRM). Each artifact is tied to a real operational problem, expressed as a deployable design (architecture, model, protocol, framework, dataset, or tool), and evaluated using simulation, adversarial testing, case studies, formal analysis, or operational metrics.
I treat identity and access management (IAM) as the control plane for trustworthy systems. The artifacts below implement identity-first security patterns—authentication, authorization, federation, and Zero Trust—across critical infrastructure, AI systems, and cyber-physical environments.
Filter by artifact type, domain, maturity, or evaluation method.
Open any artifact to view its DSRM mapping (Problem → Design → Build → Evaluate → Communicate), evaluation context, and citation formats for journals and grants.
Follow "Related Research Notes" to see the design rationale and evidence trail.
Design Science Research Methodology
Every artifact in this gallery follows the DSRM lifecycle: identifying real operational problems, designing solutions grounded in identity-first principles, building deployable artifacts, evaluating them under realistic conditions, and communicating results for reproducibility and adoption.
Showing 9 of 9 artifacts
Problem
AI agents and AI-enabled systems increasingly rely on long-lived memory, creating new attack surfaces for prompt injection persistence, data exfiltration, and policy bypass.
Artifact
A secure memory governance framework for AI-enabled systems, emphasizing policy-driven access, verification hooks, and least-privilege memory interfaces.
Problem
Post-quantum migration creates systemic risk across dependencies, certificates, protocols, and cryptographic inventories; manual transitions do not scale.
Artifact
A quantum-transition security framework for large-scale network and infrastructure environments, emphasizing governance, playbooks, and continuous verification of cryptographic posture.
Problem
V2I communications expand attack surfaces for spoofing, replay, jamming, and integrity attacks that can cascade into unsafe roadway decisions.
Artifact
An AI/ML-based V2I security protocol for anomaly detection and resilient operation, designed for transportation cyber-physical environments.
Problem
Critical infrastructure operators need detection that generalizes under novel conditions and partial observability without brittle signature dependence.
Artifact
A JEPA-informed autonomous defense framework generalized for critical infrastructure/DER environments, emphasizing predictive representations for anomaly detection and resilient control.
Problem
Scientific cyberinfrastructure faces complex trust and access challenges across heterogeneous users, workloads, and federated environments.
Artifact
A probabilistic zero-trust framework for scientific cyberinfrastructure with formal guarantees and in-situ validation orientation.
Problem
Canonical DSRM's single-pass evaluation model proves insufficient for AI systems operating in adversarial, continuously evolving environments.
Artifact
Agile-Infused Design Science Research Methodology: A systematic extension for adversarial AI systems that embeds continuous validity preservation through iterative cycles.
Problem
NIST AI RMF establishes governance principles but lacks technical enforcement mechanisms for adversarial, agentic AI systems in critical infrastructure.
Artifact
Identity-Centric Adaptive Control Core: A three-layer architecture operationalizing NIST AI RMF through A-DSRM cycles and IAM mediation for critical infrastructure.
Problem
Critical infrastructure AI faces fundamental resilience challenges under coordinated cyber-physical attacks that traditional security approaches cannot address.
Artifact
Resilience Envelopes: Dynamic operational boundaries maintained through A-DSRM iterations for AI systems in critical infrastructure across electric grids, water treatment, and transportation.
Problem
Methodological innovation requires practical tooling to achieve widespread adoption; ad-hoc approaches reduce reproducibility and increase vulnerability windows.
Artifact
Open-source toolchain implementing A-DSRM methodology with workflow engine, validity monitor, policy transformer, and experiment orchestrator for adversarial AI research.