Synchronous Scheduling Algorithms for Edge Coordinated Internet of Things

Olaniyan, Richard; Maheswaran, Muthucumaru

doi:10.1109/cfec.2018.8358725

Cited by 7 publications

(4 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The synchronization schemes proposed in [22] and [23] focused mainly on achieving synchronization without much consideration on the message overhead caused by the constant communication between worker nodes and the controller. The algorithms were based on the task attributes, time and component redundancy as well as localization of worker nodes.…”

Section: B Synchronization In Real-time Systemsmentioning

confidence: 99%

“…Unlike DSSP, where workers are expected to have the same or very similar runtimes per iteration, we assume a highly heterogeneous system where runtimes may vary from one iteration to another. The synchronization schemes in [22] and [23] incur extra communication overhead from heavy involvement of the controller. This work achieves fast synchronization by reducing the number of messages sent through clustering and by minimizing the involvement of the controller in making synchronization decisions.…”

Section: Comparison Of This Work and Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A Fast Edge-Based Synchronizer for Tasks in Real-Time Artificial Intelligence Applications

Olaniyan¹,

Maheswaran²

2020

Preprint

Self Cite

View full text Add to dashboard Cite

Real-time artificial intelligence (AI) applications mapped onto edge computing need to perform data capture, process data, and device actuation within given bounds while using the available devices. Task synchronization across the devices is an important problem that affects the timely progress of an AI application by determining the quality of the captured data, time to process the data, and the quality of actuation. In this paper, we develop a fast edge-based synchronization scheme that can time align the execution of input-output tasks as well compute tasks. The primary idea of the fast synchronizer is to cluster the devices into groups that are highly synchronized in their task executions and statically determine few synchronization points using a game-theoretic solver. The cluster of devices use a late notification protocol to select the best point among the pre-computed synchronization points to reach a time aligned task execution as quickly as possible. We evaluate the performance of our synchronization scheme using trace-driven simulations and we compare the performance with existing distributed synchronization schemes for real-time AI application tasks. We implement our synchronization scheme and compare its training accuracy and training time with other parameter server synchronization frameworks.

show abstract

Section: B Synchronization In Real-time Systemsmentioning

confidence: 99%

Section: Comparison Of This Work and Related Workmentioning

confidence: 99%

A Fast Edge-Based Synchronizer for Tasks in Real-Time Artificial Intelligence Applications

Olaniyan¹,

Maheswaran²

2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…The research problem of coordination between the cloud and the edge machines has been looked at in the work of Olaniyan and Maheswaran . The authors proposed three kinds of scheduling algorithms, ie, static, dynamic, and batch.…”

Section: Related Workmentioning

confidence: 99%

Load balancing aware scheduling algorithms for fog networks

Singh

Auluck

2019

Softw Pract Exp

View full text Add to dashboard Cite

Fog networks have attracted the attention of researchers recently. The idea is that a part of the computation of a job/application can be performed by fog devices that are located at the network edge, close to the users. Executing latency sensitive applications on the cloud may not be feasible, owing to the significant communication delay involved between the user and the cloud data center (cdc). By the time the application traverses the network and reaches the cloud data center, it might already be too late. However, fog devices, also known as mobile data centers (mdcs), are capable of executing such latency sensitive applications. In this paper, we study the problem of balancing the application load while taking account of security constraints of jobs, across various mdcs in a fog network. In case a particular mdc does not have sufficient capacity to execute a job, the job needs to be migrated to some other mdc. To this end, we propose three heuristic algorithms: minimum distance, minimum load, and minimum hop distance and load (MHDL). In addition, we also propose an ILP-based algorithm called load balancing aware scheduling ILP (LASILP) for solving the task mapping and scheduling problem. The performance of the proposed algorithms have been compared with the cloud only algorithm and another heuristic algorithm called fog-cloud-placement (FCP). Simulation results performed on real-life workload traces reveal that the MHDL heuristic performs better as compared to other scheduling policies in the fog computing environment while meeting application privacy requirements.

show abstract

“…A number of energy‐efficient, resource‐centric, and synchronization‐aware task scheduling and allocation algorithms have been proposed in the literature. Olaniyan and Maheswaran 41 proposed three task synchronization and scheduling algorithms for coordinated edge/IoT devices with stringent timing requirements. They focused not only on strict timing requirements but also the possibility of unstable connection and unavailability of nodes.…”

Section: Related Workmentioning

confidence: 99%

Real‐time partitioned scheduling: Exploiting the inter‐resource affinity for task allocation on multiprocessors

Akram

Bai

et al. 2019

Concurrency and Computation

View full text Add to dashboard Cite

Real-time edge computing is forging its place in cloud computing rapidly, and requirements for high-performance edge devices are becoming increasingly complex. Multiprocessor edge devices are an attractive choice to meet these higher performance requirements. However, multiprocessor devices encounter inherent challenges when handling on-chip shared resources. The concurrent access to these resources by the tasks requesting more than one shared resource and running on multiple processors may face huge blocking times, which can lead to missed hard real-time deadlines and cause a catastrophic system failure. To reduce the task blocking time, we propose a task allocation algorithm that takes advantage of inter-resource affinity and allocates all the tasks accessing multiple shared resources having inter-resource affinity on the same processor of the multiprocessor. The proposed algorithm reduces global resources and remote blocking, which subsequently increase schedulability of task sets and reduce the processor utilization. In our experiments, we compare the proposed task allocation algorithm SRTA with existing well-known task allocation strategies SPA, ROP, and blocking-agnostic FFD. The experimental results reveal that the SRTA, on the average, can allocate 1.80, 2.34, and 3.14 times more task sets as compared to ROP, SPA, and FFD, respectively and reduces the number of globally shared resources significantly.

show abstract

Synchronous Scheduling Algorithms for Edge Coordinated Internet of Things

Cited by 7 publications

References 16 publications

A Fast Edge-Based Synchronizer for Tasks in Real-Time Artificial Intelligence Applications

A Fast Edge-Based Synchronizer for Tasks in Real-Time Artificial Intelligence Applications

Load balancing aware scheduling algorithms for fog networks

Real‐time partitioned scheduling: Exploiting the inter‐resource affinity for task allocation on multiprocessors

Contact Info

Product

Resources

About