학술논문

Performance Evaluation of Kubernetes Distributions (K8s, K3s, KubeEdge) in an Adaptive and Federated Cloud Infrastructure for Disadvantaged Tactical Networks
Document Type
Conference
Source
2021 International Conference on Military Communication and Information Systems (ICMCIS) Military Communication and Information Systems (ICMCIS), 2021 International Conference on. :1-7 May, 2021
Subject
Communication, Networking and Broadcast Technologies
Computing and Processing
Fields, Waves and Electromagnetics
Robotics and Control Systems
Signal Processing and Analysis
Wireless communication
Performance evaluation
Military communication
Cloud computing
Adaptive systems
Computational modeling
Computer architecture
Tactical Clouds
Federated Clouds
Kubernetes
K8s
K3s
KubeEdge
Disadvantaged Tactical Networks
Network Adaptation
Language
Abstract
The tactical edge domain, primarily consisting of dismounted soldiers and vehicles on the move, are typically interconnected via wireless tactical networks that are limited in terms of bandwidth, reachability, reliability, and latency. Hence, nodes in the tactical network cannot simply rely on assured access to enterprise cloud computing. Instead, they must explore other alternative models to leverage resources that are in situ, by means of a federated cloud architecture that spans the three tiers of dismounted soldiers, vehicles on the move, and operations centers. The NATO IST-168 RTG has been exploring approaches to best exploit available resources in such a federated architecture while living within the constraints of the tactical networks. The first approach has been to evaluate Kubernetes technologies to see if they are able to be deployed over tactical networks and provide the capabilities to dynamically distribute data and computing tasks over a federated cloud infrastructure composed of multiple partner nation networks. This paper provides initial performance results for various Kubernetes distributions (K8s, K3s, KubeEdge) in federated and adaptive tactical networks, leading to recommendations for further development and experimentation.