Research Topics

Open-Source based Hybrid Cloud Infrastructure

  • VM/Container Hybrid Cloud Platform Implementation
    • OpenStack / Kubernetes
  • Multi-site Kubernetes Edge
    • StarlingX, Kube Edge
  • NFV MANO for Hybrid Cloud
    • OpenStack Tacker as MANO
  • Hybrid-cloud Infra Management
    • Monitoring Automation / RCA(Root Cause Analysis)
    • Workflow-based Zero-touch Fault Management
  • Distributed Tracing Environment Automation System
    • Proxy & Instrumentation Integrated Sidecar
  • AI-based Infra Management

Service Management in Edge-Cloud

  • NFV MANO for Kubernetes Based Cloud System
    • VNF Management Based VM and Container together Using Operator Framework
  • Container Service Migration
    • Checkpoint-base Stateful Service Migration
  • Zero-touch Service Management
    • Closed Loop Framework
  • Application Placement in Multi-Cluster Cloud
    • Policy-Based Scheduling
  • Serverless Computing Service for Edge Clouds
    • Based on Interworking Using Cross-Monitoring

Networking in Future Internet Infrastructure

  • Location/ID Separation for Edge Computing Environment
    • Service Path Optimization / Dynamic Routing
    • Service Mobility Management
  • Network Service in Cloud Native Environment
    • Service Function Chaining (SFC) Using Network Service Mesh
  • Container Networking Acceleration / Benchmarking

Hybrid cloud management [Openstack & AWS] – 5G deployment

  • Enhanced 5G network function management
    • Slicing functions and Deploying the UPF and 5G core on AWS and the RAN on OpenStack

Fault Tolerance and Auto Recovery System for High Availability over Hybrid Cloud Environment

  • Fault Tolerance and Automated Recovery Mechanisms
    • Integrating multi-cluster orchestration with fault detection
    • Workflow-based auto-recovery
  • High-availability Workload
    • Efficient conflict resolution
    • Adaptive Fault Management

FaaS Workflow-based FaaS for multi-event processing

  • Scalable and Reusable Framework
    • Accommodate Events From Multiple Sources (general cloud events and events from 5G networks)
    • Independent programming of various event combinations, conditions, workflows for different types of events

5G MEC architecture enhancements for supporting Serverless Computing

  • Computing-Aware Traffic Steering 5G Network Function to manage pre-warmed serverless deployment information at edge
  • MEC FaaS Framework enhancements to support efficient serverless auto-scaling in different edge scenarios
    • Serverless edge apps share a limited resource quota, aims to avoid over-quota resource usage
    • Serverless edge apps can be assigned any resource amount, aims to minimize resource allocation
    • Avoid serverless cold-start latency problem

Design and Implementation of Workload Migration in a Kubernetes-based Multi-Cluster Environment

  • Backup and Restore
    • First Phase: Cache-Based Quick Recovery
      (Using Distributed Cache System)
    • Second Phase: Kubernetes Resource Restoration
  • Service Continuity
    • Utilizing Transaction Queues after Cache-Based Restoration
    • Reducing Service Downtime with Fast Initial Restore
  • Failover
    • Automated Workload Migration

Automated Resource Quota Scaling in Multi-Cluster Kubernetes Environments

  • Dual-Method Solution for Resource Quota Scaling in Multi-Cluster Environments
    • Reactive Scaling
    • Proactive Scaling