Adaptive Real-Time Scheduling Algorithms for Embedded Systems

TL;DR

This paper surveys adaptive real-time scheduling algorithms for embedded systems, focusing on feedback control, task interdependence, predictive methods, and power management to handle dynamic environments and safety-critical requirements.

Contribution

It provides a concise overview of key mechanisms, trade-offs, evaluation metrics, and ongoing challenges in adaptive real-time scheduling for embedded systems.

Findings

Summarizes feedback-based control and interdependence models.

Highlights trade-offs between adaptivity and predictability.

Discusses evaluation metrics and safety considerations.

Abstract

Embedded systems are becoming more in demand to work in dynamic and uncertain environments, and being confined to the strong requirements of real-time. Conventional static scheduling models usually cannot cope with runtime modification in workload, resource availability, or system updates. This brief survey covers the area of feedback-based control (e.g., Feedback Control Scheduling) and interdependence between tasks (e.g., Symbiotic Scheduling of Periodic Tasks) models. It also borders on predictive methods and power management, combining methods based on Dynamic Voltage and Frequency Scaling (DVFS). In this paper, key mechanisms are briefly summarized, influencing trade-offs relating to adaptivity/predictability, typical metrics of evaluation, and ongoing problems, especially in situations where safety is a critical factor, giving a succinct and easy-to-understand introduction to researchers and practitioners who have to cope with the changing environment of adaptive real-time systems.