Punctual Cloud: Achieving Punctuality for Time-Critical Cloud Control Systems

Abstract

Cloud Control Systems (CCSs) harness the power of cloud resources to carry out intense computational tasks, however, they face challenges in delivering time-critical control signals due to network and cloud-induced delays. In this paper, we introduce a novel framework, "Punctual Cloud", designed to enhance the timely delivery of control signals in CCS. This framework ensures that control signals are calculated in the cloud and delivered promptly, minimizing system-induced delays. The punctual cloud framework allows a basic controller to manage dynamic, time-sensitive processes over the cloud and is implemented in a microservice architecture to simplify CCS deployment and maintenance. Our evaluation indicates that our framework, while requiring less computational effort, is capable of tolerating longer delays compared to an optimal controller like Model Predictive Control (MPC), which is typically used in time-delay control systems.