Research output
Grouped by year. Click a paper or badge to open the manuscript or IEEE record.
Transformer-based deep reinforcement learning scheduler for service provisioning in cloud-edge systems — improving SLA satisfaction and latency under volatile workloads.
IEEE Transactions on Cloud Computing
A zero-touch, DRL-driven autoscaling framework for cloud-native Kubernetes edge environments, enabling adaptive orchestration beyond heuristic HPA policies.
IEEE GLOBECOM 2025 · Taipei, Taiwan
Distributed, cloud-native MARL framework for 6G that uses Dueling Double Deep Q-Networks to optimize containerized network function placement and migration.
IEEE ICC 2025 · Montreal, Canada
SLA-aware autoscaling for Kubernetes combining online RL-based scaling with an offline RLOps training pipeline.
IEEE XploreIEEE NFV-SDN 2025 · Athens, Greece
IEEE CAMAD 2024 · Athens, Greece