ACM SIGBED Review

In designing and analyzing real-time systems, the central problem resides in finding a feasible schedule for a given task set, if one exists. A lot of research effort has been carried out in approaching the various aspects of task scheduling. While most results have been achieved for preemptive scheduling, the non-preemptive case has still room for improvement, due to its complexity. In addition, the widespread usage of cyber-physical systems (CPS) is putting real-time scheduling in the position to deal with new challenges, as additional (and sometimes particular) requirements are raised by such systems. This paper, which continues the previous work of the authors, introduces a new lower bound on the number of processing units that allows a feasible schedule of a task set for both preemptive and non-preemptive scheduling. As a specific CPS issue, the necessity of moving the processing units in space, which incurs additional time requirements, is considered. The single-instance case is first discussed, then the results are extended to the periodic case.

Vlad Rădulescu

Ştefan Andrei

Albert M. K. Cheng

Bounding execution resources for the task scheduling problem in cyber-physical systems

网络物理系统中任务调度问题的执行资源约束

Android is becoming used more frequently in domains that expect some real-time guarantees. To facilitate the adoption of Android for programming real-time systems, this work presents a first effort to apply real-time scheduling theories to a real-time extension on Android, RTDroid. We integrate real-time properties specified in RT-Droid's application manifest with an existing real-time scheduling framework, Cheddar. We leverage Cheddar to perform schedulability analysis and feasibility tests, based on the properties of RTDroid application components specified in the application manifest. This paper details our integration process and reports our experience of validating real-time properties in a real-time application developed on RTDroid.

Yin Yan

Lukasz Ziarek

Application validation on RTDroid

RTDroid的应用验证

Abort-and-Restart model is used in Priority-based Functional Reactive Programming in which higher priority tasks can preempt lower priority tasks and the lower priority tasks are aborted and restarted after the higher priority tasks have finished execution. This paper discusses a potential of improving schedulability in P-FRP systems by using preemption threshold that it allows a task to only disable preemption of tasks up to a specified threshold priority. Also, a sufficient schedulability test condition is studied in this paper for P-FRP tasks using preemption threshold, which is a critical problem to be solved in order to explore the potential benefit.

Jian (Denny) Lin

P-FRP task scheduling with preemption threshold

P-FRP任务调度与抢占阈值

The current era can be characterized by the massive reliance on computing platforms in almost all domains, such as manufacturing, defense, healthcare, government. However, with the increased productivity, flexibility, and effectiveness that computers provide, comes the vulnerability to cyber-attacks where software, or even firmware, gets subtly modified by a hacker. The integration of a Trusted Platform Module (TPM) opts to tackle this issue by aiding in the detection of unauthorized modifications so that devices get remediation as needed. Nonetheless, the use of a TPM is impractical for resource-constrained devices due to power, space and cost limitations. With the recent proliferation of miniaturized devices along with the push towards the Internet-of Things (IoT) there is a need for a lightweight and practical alternative to the TPM. This paper proposes a cost-effective solution that incorporates modest amounts of integrated roots-of-trust logic and supports attestation of the integrity of the device's boot-up state. Our solution leverages crypto-acceleration modules found on many microprocessor and microcontroller based IoT devices nowadays, and introduces little additional overhead. The basic concepts have been validated through implementation on an SoC with an FPGA and a hard microcontroller. We report the validation results and highlight the involved tradeoffs.

Tom Broström

John Zhu

Ryan Robucci

Mohamed Younis

IoT boot integrity measuring and reporting

物联网启动完整性测量与报告

For the Internet of Things (IoT) applications that send a few bytes of sensor information infrequently, several long-range IoT technologies have been conceived. Narrowband IoT (NB-IoT) is one of them that stands out due to its extended coverage, high penetrability, and high reliability features. Among the primary goals set for the standard, low device power consumption is significant as the devices are typically deployed with batteries. In this paper, we characterize an NB-IoT device with respect to its energy consumption in different coverage classes for various scenarios. Based on the measurements, we estimate the lifetime of the device. We find that the lifetime of the device is significantly affected by the coverage class of the device. In order to augment the battery, we investigate the possibility of using ambient energy sources in the context of a smart home. For this we analyze real-world data sets, and find that the harvesting technologies can increase the average lifetime of the device. However, due to the spatio-temporal variations in the amount of energy that can be harvested, there are several challenges to be dealt with. We list the keys challenges of using energy-harvesting in NB-IoT, and those inherent to NB-IoT.

A. Haridas

V. S. Rao

R. V. Prasad

C. Sarkar

Opportunities and challenges in using energy-harvesting for NB-IoT

利用能量采集技术在NB-IoT中的机遇与挑战

Virtualization is the main building block of many architectures and systems from embedded computing to large scale data-centers. Managing efficiently computing resources and their power consumption becomes fundamental to optimize the performance of the workloads running on those systems, however, hardware tools like Intel RAPL can only introduce power caps without considering performance. This paper presents a performance-aware power capping orchestrator for the Xen hypervisor. The tool exploits hybrid power management techniques to minimize power consumption respecting a given SLA, leveraging RAPL and overcoming its limitations. Experimental evaluation shows that the proposed approach guarantees good results for almost all the analyzed benchmarks (e.g. CPU-, memory- and IO-bound).

Rolando Brondolin

Marco Arnaboldi

Marco D. Santambrogio

Power consumption management under a low-level performance constraint in the Xen hypervisor

Xen虚拟机管理程序中低性能约束下的功耗管理

Intermittently-powered embedded devices (IPDs) are getting wide-spread attention these days. However, running periodic real-time tasks on these devices remains a challenging problem due to the lack of support for data freshness guarantees, timekeeping, and schedulability analysis. Especially, while many sensing tasks require long atomic operations for data acquisition from sensors, most prior work on IPDs assumes compute-only workloads and disregards such sensor operations. In this paper, we present a new energy scheduling scheme to execute periodic real-time tasks with atomic sensing operations. Our scheme keeps track of time and ensures the periodic execution of sensing tasks while efficiently utilizing intermittent power sources. We provide schedulability analysis to determine if a task is schedulable in a given charging setup, and extend this idea for scheduling multiple tasks. As a proof-of-concept, we design a custom programmable RFID tag device, called R'tag, and demonstrate the effectiveness of our proposed techniques in a realistic sensing application. We compare the baseline approach and the proposed scheme in both simulation and real platforms. Experimental results show that the proposed method outperforms the baseline approach in terms of task scheduling, timekeeping, and periodic sensing.

Mohsen Karimi

Hyoseung Kim

Energy scheduling for task execution on intermittently-powered devices

间歇供电设备上的任务执行能量调度

Hands-on experience is crucial to truly understanding the principles of operating systems (OS). This paper describes a preemptive, multithreaded embedded OS assigned as a project in a junior-level OS course. It targets the popular 8051 instruction set architecture (ISA) and can run with as little as 128 bytes of RAM and 1 KB of program memory with preemption and synchronization primitives such as semaphores. It is written in C and assembly and compiled using the open-source Small Device C Compiler (SDCC). It runs on the free EdSim51 simulator, which simulates common peripherals such as LCD, keypads, 7-segment LEDs, ADC, and UART all in high fidelity. This course project has received positive feedback from students who took the course.

Wei-Fang Sun

Shih-Hsien Chan

Zhe-Ting Liu

Yu-Hung Yeh

Pai H. Chou

Quatros

四元组

ARM big.LITTLE architectures are spreading more and more in the mobile world thanks to their power-saving capabilities due to the use of two ISA-compatible islands, one focusing on energy efficiency and the other one on computational power. This architecture makes the problem of energy-aware task scheduling particularly challenging, due to the number of variables to take into account and the need for having lightweight mechanisms that can be readily computed in an operating system kernel scheduler. This paper presents a novel task scheduler for big.LITTLE platforms, combining the well-known Constant Bandwidth Server algorithm with a power-aware per-job migration policy. This achieves real-time adaptation of the CPU islands' frequencies based on the individual cores' overall utilization, as available in the scheduler thanks to the use of the resource reservation paradigm. Preliminary results obtained by simulations based on modifications to the open-source RTSim tool show that the proposed technique is able to achieve interesting performance/energy trade-offs.

Agostino Mascitti

Tommaso Cucinotta

Mauro Marinoni

An adaptive, utilization-based approach to schedule real-time tasks for ARM big.LITTLE architectures

一种针对ARM big.LITTLE架构的实时任务调度自适应利用率方法

As system complexity grows, embedded operating systems increasingly face the challenge to adhere to various non-functional constraints, such as response times and power limits. These requirements sometimes contradict and, often, no solution satisfies all constraints under all conditions. Changes in environmental conditions, application-level requirements, and user response time expectation hence demand for system-wide adaptions to resource management. We find that process synchronization constitutes a simple yet effective leverage point to balance between timing-related and energy-related constraints. This paper presents Earl, an implementation of reconfigurable locks in Linux that enables seamless transitions between high-performance and low-power operating modes.

ISSN	1551-3688
h-index	11

涉及主题	计算机科学工程类操作系统程序设计语言数学嵌入式系统经济分布式计算计算机网络物理电信管理电气工程生物量子力学数学优化运营管理实时计算地理人工智能计算机安全调度（生产过程）热力学
出版信息	出版商: Association for Computing Machinery (ACM)，出版周期: ，期刊类型: journal，开源期刊: 否
基本数据	创刊年份: 2018，原创研究文献占比: 0.00%，自引率:N.A.， Gold OA占比: 0.00%

ACM SIGBED Review

期刊引文格式

有 1 位以上作者的期刊

有 2 位作者的期刊

有 3 位作者的期刊

有 5 位以上作者的期刊

书籍引用格式

学位论文引用格式

网页引用格式

专利引用格式

手工熬夜修改参考文献？研飞自动匹配期刊，一键轻松成稿，支持 Word/WPS

投稿经验分享