I work on mobile edge computing, task offloading, distributed scheduling, edge intelligence, and reproducible machine-learning systems.
My current projects connect optimization-based MEC research with reliable distributed execution infrastructure.
- Mobile Edge Computing
- Task Offloading and Resource Allocation
- Distributed Task Scheduling
- Edge Intelligence and ML Systems
- Reproducible Optimization and Evaluation
Research artefact for QoE- and fairness-aware MEC task-offloading optimization. It organizes configuration-driven experiments, paired repeated runs, baseline comparison, ablation, sensitivity and scalability studies, statistical analysis, and raw reproducibility artefacts. Manuscript in preparation.
Redis-backed distributed task execution with a FastAPI control plane, atomic task claiming, multiple workers, lifecycle controls, heartbeats, retries, timeouts, cooperative cancellation, observability, Docker Compose, tests, and policy benchmarks. It is a research-oriented distributed-systems implementation and foundation for later MEC and ML-scheduling experiments.
A lightweight, deterministic MEC baseline simulator for comparing local, edge, and cloud execution under supplied configurations and seeds. It is not a production scheduler or a real deployment.
A reproducible two-week supervised-learning study project covering leakage-free evaluation, model selection, and an IoT predictive-maintenance exercise. It documents learning and engineering practice rather than a novel research contribution.
- Connect MEC optimization with online distributed scheduling.
- Evaluate scheduling under dynamic workloads with reproducible system benchmarks.
- Study learning-based scheduling after the foundational system is stable.
- Improve generalization, robustness, and fairness evaluation.
Programming: Python, Java, C Systems: FastAPI, Redis, Docker Compose, REST APIs, concurrent workers ML and Research: scikit-learn, optimization, statistical evaluation, reproducible experiments