Elsevier

Computer Communications

Volume 158, 15 May 2020, Pages 73-84
Computer Communications

Resource scheduling for piano teaching system of internet of things based on mobile edge computing

https://doi.org/10.1016/j.comcom.2020.04.056Get rights and content

Abstract

The effective operation of the piano teaching system of the Internet of Things requires the effective support of virtualization technology. In particular, on the basis that the edge computing standards and systems are not yet mature, the resource scheduling problem of edge computing needs to be studied from the actual point of view. In order to improve the effective operation of the piano teaching system of Internet of Things, this study analyzes the resource scheduling of delay-sensitive applications, sets the resource scheduling mode based on the space–time difference of the edge container load in a multi-cluster environment, and proposes a cross-cluster scheduling strategy. Simultaneously, this study uses simulation experiments to analyze the performance of the strategy proposed in this paper. The research results show that the strategy proposed in this paper can perform delay-insensitive application scheduling during system operation, achieve multi-cluster collaborative scheduling goals, and make the load between clusters more balanced.

Introduction

The Internet of Things is a core component of the intelligent teaching information platform. It uses advanced sensing technology, network technology, computing technology, control technology, and intelligent technology to comprehensively perceive the teaching process through cameras and other devices. Moreover, it aggregates its collected environment and status information, etc., and conducts large-scale, large-capacity data transmission and interaction between multiple systems to analyze and process the teaching information in a timely and sufficient manner, thereby supporting full control of each student, and ultimately effectively improving teaching efficiency.

In the past ten years, mobile cloud computing (MCC) technology has experienced rapid development. Due to high reliability, good scalability, and strong computing power, the technology is used as the basic support for mobile applications, from leading giants to startup small and micro enterprises. For example, the professional cloud service of international e-commerce giant Amazon (Amazon Web Services, AWS) has provided cloud platform support to thousands of enterprises in more than 190 countries and regions by deploying data centers around the world and more than 2 million cloud servers. Nowadays, on the new front of the industry 4.0 era, mobile applications are facing increasingly stringent service quality requirements [1] (Quality-of-Service, QoS). Moreover, the era of big data 2.0 will fuel the blowout growth of massive data [2]. In addition, the maturity and implementation of 5G bearer technology will inevitably make the core network more severely tested [3]. In this situation, MCCs that provide centralized services will likely be trapped, and it is difficult to achieve the ambitious vision of a millisecond response [4]: On the one hand, real-time communication is difficult to achieve due to network congestion caused by massive requests rushing to data centers. On the other hand, the heavy load of the data center makes it very easy to become a service bottleneck that delays the response. What is more, the high expansion cost of data centers is obviously not suitable for the exponentially rising demand scale. Therefore, the design needs of the next-generation cloud service system are imminent and eager, which has stimulated the enthusiasm of researchers at home and abroad [5]. In recent years, a new computing method called Mobile Edge Computing (MEC) has dawned and is expected to take on the responsibility of the next generation of cloud service systems. The European Telecommunications Standardization Organization defines MEC as “a cloud platform that provides services directly in the wireless access network close to users and provides high-bandwidth, low-latency computing and storage services to mobile users near data sources” [6]. In the future plan of MEC, the central cloud function of MCC will be “submerged” to tens of billions of network edge devices, making it “edge cloud server”. These servers not only have the basic network functions of network access and forwarding traffic, but also undertake hosting services and cloud computing functions for processing requests and provide mobile users with “edge cloud services” that can be enjoyed with a single hop network distance [7]. With the blessing of 5G bearer technology, MEC will definitely show great promise in the fields of emerging mobile applications, smart cities, industrial Internet of things, and smart homes [8]. From the perspective of the network structure, the MEC service system located near the terminal access network not only offloads the workload for the MCC data center, relieves network congestion, and improves service quality for the MCC data center, but also makes up the shortcomings of the MCC in computing response speed and content distribution efficiency, and improves the user experience. Compared with substitution and innovation, MEC is complementary to MCC and has complementary advantages, and the two will work together to create a rich and diverse cloud service ecosystem.

Today, the “second half” of the mobile Internet industry has begun: Pervasive computing is in the ascendant, the Internet of Everything is beginning to flourish, artificial intelligence is about to fall, and complex and emerging applications are increasingly mobile. What is​ more, resource-intensive, delay-sensitive computing tasks such as view processing, big data technology, and reinforcement learning have imposed severe requirements on computing device storage, storage, and battery life. However, consumer-grade mobile terminals are limited by the shortage of computing power resources and it is difficult to meet high standards of service alone.

In order to resolve the contradiction between increasing application requirements and limited resources, an effective solution is to enable users to submit some or all of the computing tasks to resource-rich cloud servers for processing. The aforementioned delegated execution process is called Computing Offloading. In the near future, due to its deployment advantages and geographical advantages, MEC is bound to become a mainstay in providing computing offload services. Computing offloading technology has important value: For mobile users, computing offloading technology expands the computing capabilities of mobile devices, saves energy consumption costs, extends its running time, and improves user experience. For operators and service providers, computing offloading technology makes full use of network resources, improves production efficiency, and can create huge benefits. In the field of MEC research, computational offloading technology has received much attention due to its higher research value. To sum up, the research topic in this paper is both practical and theoretical. Moreover, this study conducts research and analysis on resource scheduling of the piano teaching system of the Internet of Things based on mobile edge computing.

Section snippets

Related work

Mobile Cloud Computing (MCC), as an example of the organic combination of communication networks and the Internet, is the most effective way to provide mobile users with cloud computing [9]. Typical business applications are such as Apple iCloud and Microsoft OneDrive. MCC service providers provide users with cloud services through a centrally operated data center, and allow users to lease resources on demand, dynamically deploy applications, and enjoy computing services and storage resources 

Cluster edge cloud resource scheduling strategy

For multi-cluster edge clouds, there are two major technical points that need attention: one is resource allocation, and the other is load balancing. This chapter will focus on the multi-cluster edge cloud framework, and respectively introduce the scheduling strategies of delay-sensitive applications and delay-insensitive applications in the framework to effectively implement resource allocation and solve load balancing problems.

Delay-insensitive application scheduling strategy

Delay-insensitive applications have lower priority in the cluster than delay-sensitive applications. When the cluster has limited resources, it will release the resources occupied by delay-insensitive applications, prioritize the resource requirements of delay-sensitive applications, and schedule delay-insensitive applications into other clusters. Delay-insensitive application scheduling is divided into three sub-steps: trigger scheduling, application selection, and cluster selection. This

Algorithm simulation tests

System testing is mainly divided into the effectiveness and performance analysis of the scheduling strategy of delay-sensitive applications and the effectiveness and performance analysis of the scheduling strategy of delay-insensitive applications. The system test hardware environment includes a PC, a DELLPowerEdge 820 server, two DELLPowerEdge R730 servers, and a Tencent cloud server. On the server, a KVM virtual machine is created and a Kubernetes cluster is set up using the virtual machine.

Conclusion

In order to improve the effective operation of the piano teaching system of the Internet of Things, this study analyzes the resource scheduling of delay-sensitive applications and proposes a multi-cluster edge cloud framework based on application-sensitive delay differences. The framework distinguishes application categories based on domain names and performs multi-cluster management through a master–slave model. On this basis, this paper studies the framework’s deployment architecture,

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Funded projects

2013 Hunan Philosophy and Social Science Foundation, China Project Name of Project “Ecological Research on Xiangxi Flower Lantern Play in the Perspective of Intangible Culture” Project Number: 13YBB173.

References (34)

  • XiaohengD. et al.

    Integrated trust based resource cooperation in edge computing

    J. Comput. Res. Dev.

    (2018)
  • YangJ. et al.

    Marine depth mapping algorithm based on the edge computing in Internet of things

    J. Parallel Distrib. Comput.

    (2018)
  • PengY. et al.

    Design and modeling of survivable network planning for software-defined data center networks in smart city

    Int. J. Commun. Syst.

    (2018)
  • MohanN. et al.

    Edge-Fog Cloud: A Distributed Cloud for Internet of Things Computations

    (2017)
  • ParkD. et al.

    LiReD: A light-weight real-time fault detection system for edge computing using LSTM recurrent neural networks

    Sensors

    (2018)
  • FanQ. et al.

    Application aware workload allocation for edge computing based IoT

    IEEE Internet Things J.

    (2018)
  • AbbasN. et al.

    Mobile edge computing: A survey

    IEEE Internet Things J.

    (2017)
  • Cited by (0)

    View full text