Links
Research Interests & Directions
-
Real-time Serverless and FaaS Scheduling
Designing latency-bounded serverless platforms; schedulability analysis and heuristic placement across multi-node FaaS clusters to guarantee end-to-end deadlines and SLOs. -
State Management for Cloud-Native Applications
Minimizing state access latency and tail behavior via data locality, replication, and consistent partitioning strategies in microservices and event-driven systems. -
Edge/MEC Platforms: Cost–Latency Optimization
Joint placement and scaling of services at the edge to meet latency constraints while optimizing operational cost, leveraging SDN/NFV-enabled architectures. -
Software-Defined Networking and Hybrid Migration
Performance analysis and transition strategies from legacy to SDN-based infrastructures, focusing on reliability, hybrid topologies, and operational resilience. -
Cloud-Native Reliability and Failure Prediction
Lightweight observability pipelines and data-driven methods to predict application failures and degradations in cloud-native stacks, enabling proactive mitigation. -
QoE-Aware Resource Management
Quality-of-experience modeling for latency-sensitive services (e.g., interactive video, conferencing), with telemetry-driven resource allocation and adaptive control. -
Function Composition and Fusion in Serverless
Optimizing end-to-end latency and cost by fusing dependent functions and co-locating data/compute, supported by theoretical models and systems validation. -
Deterministic Edge for Industrial/AR Use Cases
Integrating deterministic local clouds and time-sensitive networking for industrial control and augmented reality, with overload-aware latency management. -
Network Slicing and QoS Engineering
Placement and resource governance for network slices under QoS/QoE objectives, using multi-objective optimization across access and transport layers. -
Academia–Industry Co-development
Translating research into practice through industrial collaborations (e.g., telecom), focusing on observability, reliability, and real-time guarantees in production.