Timing Analysis of Asynchronized Distributed Cause-Effect Chains

Günzel, Mario; Chen, Kuan-Hsun; Ueter, Niklas; Brüggen, Georg von der; Dürr, Marco; Chen, Jian-Jia

doi:10.1109/rtas52030.2021.00012

Cited by 25 publications

(4 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This ensures that a single job invocation of a task does not overwrite its already written output in an asynchronous communication. Moreover, the fixed execution time tightly bounds a pipeline's end-to-end latency [32].…”

Section: Scheduling Modelmentioning

confidence: 99%

“…They use forward and backward cause-effect chains to derive a tighter upper bound on E2E delay. After converting the sporadic task model to the periodic task model as done in a subsequent work [32], the worst-case E2E delay (also called the maximum reaction time) of a pipeline considering all the tasks are released at the critical instant [48] (initial release offset is 0), as proved by Dürr et al, is the following:…”

Section: End-to-end Delay (E) Computationmentioning

confidence: 99%

“…There are other research studies that explore the end-to-end timing analysis of task chains [6,32,40,41] in practical scenarios such as in drones [16], and in ROS [13,65]. Proposed scheduling algorithms based on these approaches rely on job release times [7,17].…”

Section: Related Workmentioning

confidence: 99%

“…Task pipelines, or cause-effect chains, have received increased attention in recent research work [17,20,32,40,41,60], partly because of their usage in well-known software packages such as ROS [13,65]. Although real-time task pipelines have long been studied [23,49,50], the application of constraints on a pipeline has received little attention [19].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

[Solution] End-to-end Scheduling of Real-time Task Pipelines on Multiprocessors

Sinha¹,

West²

2022

Journal of Systems Research

View full text Add to dashboard Cite

Real-time embedded systems require significant analyses to ensure that their timing and functional correctness are met. With increasing complexities in hardware and software, this is a challenging proposition. Researchers have studied task-pipelines (also known as cause-effect chains) to deal with new applications in domains such as IoT and the advent of software frameworks such as ROS. In spite of increased usage of task pipelines there is scant work in analysis tools that can guarantee end-to-end properties. This paper aims to fix that by developing a constraint solver algorithm, "CoPi", that can derive runtime budgets and periods of individual pipelines tasks. This information can then be used to analyse the end-to-end constraints of the real-time system.The paper then extends this to multiprocessor systems that can handle multiple task pipelines, which in itself is quite a challenge. This fills a gap in the scheduling literature of pipeline-based task models and presents a nice heuristic optimization. The authors also present an extensive evaluation on open-source baselines for a large parameter space. Given the QoS constraints in many datacenter applications today, the paper also highlights possible design choices for scheduling in datacenter systems, thus demonstrating wide applicability. Hence CoPi has the potential to be influential in real-time, IoT and even datacenter systems.

show abstract

Section: Scheduling Modelmentioning

confidence: 99%

Section: End-to-end Delay (E) Computationmentioning

confidence: 99%