Edge server placement in mobile edge computing
Introduction
In recent years, mobile smart devices have become increasingly important as a tool for entertainment, learning, news, businesses, and social networking for smarter living [[3], [28]]. The next-generation mobile networks aim to accelerate the development of smart cities, by not only increasing the data delivery rates but also increasing the amount of infrastructures used by smart city services and applications [28]. Although mobile applications are emerging and becoming computation-intensive, the computing capacity of mobile devices remains limited owing to the resource constraints of mobile devices (e.g., processing power, battery lifetime, and storage capacity), such that mobile users do not receive the same satisfaction compared to desktop device users [13]. An effective approach to enhancing the performance of mobile applications is to offload some of their tasks to remote resource-rich clouds [[2], [20], [9], [11]].
However, the cloud is often remotely located and far from mobile users, and the data transfer delays between users and the cloud can be long and unpredictable. This is especially undesirable for mobile applications in which an immediate response time is critical to users, such as reality-augmenting applications and mobile multiplayer gaming systems. To overcome the above-mentioned problem, cloudlet-based offloading has been proposed, where mobile devices offload computational process to a computing infrastructure (i.e., cloudlet) that is in relatively close proximity to the users using Wi-Fi access points (APs) [25]. Compared to cloud computing, cloudlets are less effective for the following reasons: cloudlets can be accessed only by a Wi-Fi AP, which covers only small regions, and cloudlets are less resourceful compared to the cloud, so they are not scalable in terms of service and resource provisioning.
To overcome the above challenges, mobile edge computing is proposed [[21], [26]]. Mobile edge computing enables mobile users to access IT and cloud computing services in close proximity within the range of radio access networks [[24], [1]]. This approach enables the computation and storage capacity from the core network to be transferred to the edge network in order to reduce latency. Edge servers can be deployed in close proximity to enable devices to offload some of their mobile application workload to realize significant improvements in the quality of mobile user experiences.
Most existing studies have focused on offloading the workloads of mobile users to cloudlets to enable mobile devices to realize energy savings, and this approach assumes that the cloudlets have already been placed [[7], [10], [15], [16]]. Little attention has been paid to the effect of offloading the workloads of mobile users to edge servers and the placement of edge servers on the performance of mobile applications. In this paper, we focus on the edge server placement in a mobile edge computing environment that provides wireless internet coverage for mobile users in a large-scale metropolitan area. First, a large number of mobile users access edge servers in mobile edge computing environments because the metropolitan area that it covers has a high population density [27]. Secondly, because of the size of the network, service providers can take advantage of economies of scale when offering edge server services by making edge server services more affordable to the general public.
However, the placement of edge servers in mobile edge computing environments is challenging. The locations of edge servers are critical to the access delays of mobile users and the resource utilization of edge servers, especially in smart cities that include several hundreds or thousands of base stations through which mobile users access the edge servers. Owing to the large size of these networks, inefficient edge server placement will result in long access delays and heavily unbalanced workloads among edge servers, i.e., some of the edge servers will be overloaded while others are underutilized, or even idle. Therefore, the strategic placement of edge servers will significantly improve the performance of various mobile applications such as edge server access delay.
We assume that each edge server has the same limited computing resource to process mobile user requests, i.e., in this study, each edge server is identical, and edge servers are placed at some base station locations for mobile user access. The objective is to balance the workload among edge servers and minimize the edge server access delay. The challenge associated with such placements is determining (1) the locations in which edge servers should be placed, and (2) which base stations should be assigned to which edge servers, which we show to be an NP-hard problem. The main contributions of this paper are as follows:
(1) We formulate the edge server placement problem as a multi-objective constraint optimization problem.
(2) We adopt mixed integer programming (MIP) to find the optimal edge server placement with workload balancing among edge servers and minimizing the edge server access delay.
(3) Experimental results that are based on datasets from about 3000 of the base stations operated by Shanghai Telecom show that our approach outperforms other approaches.
The rest of the paper is organized as follows: In Section 2, we review related work. In Section 3, we introduce the system model and problem definitions, and we propose our edge server placement approach. In Section 4, we illustrate the comparative experimental evaluation results, while in Section 5, we conclude the paper, including an outlook on our future work.
Section snippets
Related work
Few studies have focused on edge server placement in mobile edge computing environments. To the best of our knowledge, this is the first study to consider the placement of edge servers in a mobile edge computing environment. However, in recent years, there have been many works on cloudlet placement [[14], [33], [32]]. Cloudlets are typically described as computers that are deployed at Wi-Fi APs in a network and act to offload the destinations of mobile users [[31], [25], [8]]. In the mobile
Our approach
In Section 3.1, we first introduce the system model and define some notations that are used throughout the rest of the paper. Then, we present the problem definition in Section 3.2. In Section 3.3, we formalize the edge server placement problem as a multi-objective optimization problem, and we describe the edge server placement model. Finally, in Section 3.4, we adopt MIP to find the optimal solution in terms of workload balancing and access delay. Related notations are explained in detail in
Performance evaluation
In this section, we implement our proposed approach to verify the performance, and we evaluate our proposed approach compared with several representative placement approaches in terms of workload balancing, access delay under various edge server workloads, and the placement of different numbers of edge servers. Results of extensive experimental evaluations show that our proposed approach is both effective and efficient.
Conclusion
Edge computing is an important emerging technology that can be used to extend the computation and storage capabilities by offloading processing workload from the cloud in order to reduce mobile edge computing network latency on mobile devices. In this study, we first investigate the edge server placement problem in a large-scale mobile edge computing environment, with the objective of balancing the workload between edge servers and minimizing the edge server access delay. Then, we formulate the
Acknowledgment
This work was supported in part by the National Science Foundation of China (Grant No. 61472047).
Shangguang Wang is an associate professor at the State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications (BUPT). He received his Ph.D. degree at BUPT in 2011. He is Vice Chair of IEEE Computer Society Technical Committee on Services Computing, President of the Service Society Young Scientist Forum in China and served as General Chair of Collaborate-Com 2016, General Chair of ICCSA 2016, TPC Chair of IOV 2014, and TPC Chair of SC2 2014.
References (33)
- et al.
Network-centric performance analysis of runtime application migration in mobile cloud computing
Simul. Model. Pract. Theory
(2015) - et al.
Application optimization in mobile cloud computing: Motivation,, taxonomies and open challenges
J. Netw. Comput. Appl.
(2015) - et al.
Application partitioning algorithms in mobile cloud computing: Taxonomy review, and future directions
J. Netw. Comput. Appl.
(2015) - A. Ahmed, E. Ahmed, A survey on mobile edge computing, in: 2016 Proceedings of the International Conference on...
- A. Ceselli, M. Premoli, S. Secci, Cloudlet network design optimization, in: 2015 Proceedings of the Interbational...
- et al.
A constant-factor approximation algorithm for the k-median problem
- et al.
Efficient multi-user computation offloading for mobile-edge cloud computing
IEEE/ACM Trans. Netw.
(2016) - et al.
Clonecloud: elastic execution between mobile device and cloud
- S. Clinch, J. Harkes, A. Friday, N. Davies, M. Satyanarayanan, How close is close enough? Understanding the role of...
- A.E.C. Cloud, Amazon web services, Retrieved November, 9 (2011)...
MAUI: making smartphones last longer with code offload
Cloud computing and SaaS as new computing platforms
Commun. ACM
Dynamic VMs placement for energy efficiency by PSO in cloud computing
J. Exp. Theor. Artif. Intell.
A survey of mobile cloud computing: architecture, applications, and approaches
Wirel. Commun. Mob. Comput.
Optimal cloudlet placement and user to cloudlet allocation in wireless metropolitan area networks
IEEE Trans. Cloud Comput.
Cuckoo: a computation offloading framework for smartphones
Cited by (341)
Server placement in mobile cloud computing: A comprehensive survey for edge computing, fog computing and cloudlet
2024, Computer Science ReviewPrivacy-preserving offloading scheme in multi-access mobile edge computing based on MADRL
2024, Journal of Parallel and Distributed ComputingEnhanced multi-objective gorilla troops optimizer for real-time multi-user dependent tasks offloading in edge-cloud computing
2023, Journal of Network and Computer ApplicationsLarge-scale mobile users deployment optimization based on a two-stage hybrid global HS-DE algorithm in multi-UAV-enabled mobile edge computing
2023, Engineering Applications of Artificial IntelligenceEnergy-aware server placement in mobile edge computing using trees social relations optimization algorithm
2024, Journal of SupercomputingMulti-layer edge resource placement optimization for factories
2024, Journal of Intelligent Manufacturing
Shangguang Wang is an associate professor at the State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications (BUPT). He received his Ph.D. degree at BUPT in 2011. He is Vice Chair of IEEE Computer Society Technical Committee on Services Computing, President of the Service Society Young Scientist Forum in China and served as General Chair of Collaborate-Com 2016, General Chair of ICCSA 2016, TPC Chair of IOV 2014, and TPC Chair of SC2 2014.
Yali Zhao received bachelor’s degree in computer science and technology from Shandong University, in 2013. Currently, she is a Master Degree Candidate at the State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications. Her research interests include service computing and edge computing.
Jinliang Xu received the bachelor’s degree in electronic information science and technology from Beijing University of Posts and Telecommunications in 2014. Currently, he is a Ph.D. candidate in computer science at the State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications. His research interests include Service Computing, Information Retrieval, and Crowdsourcing.
Jie Yuan received a Bachelor of Engineering degree in computer science and technology from Xi’an Jiaotong University, in 2016. Currently, he is a master of candidate at the State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications. His research interests include mobile edge computing.
Ching-Hsien Hsu is a professor and the chairman in the CSIE department at Chung Hua University, Taiwan; He was distinguished chair professor at Tianjin University of Technology, China, during 2012–2016. His research includes high performance computing, cloud computing, parallel and distributed systems, big data analytics, ubiquitous/pervasive computing and intelligence. He has published 100 papers in top journals such as IEEE TPDS, IEEE TSC, IEEE TCC, IEEE TETC, IEEE System, IEEE Network, ACM TOMM and book chapters in these areas. Dr. Hsu is serving as editorial board for a number of prestigious journals, including IEEE TSC, IEEE TCC. He has been acting as an author/co-author or an editor/co-editor of 10 books from Elsevier, Springer, IGI Global, World Scientific and McGraw-Hill. Dr. Hsu was awarded nine times distinguished award for excellence in research from Chung Hua University. He is vice chair of IEEE TCCLD, executive committee of IEEE TCSC, Taiwan Association of Cloud Computing and an IEEE senior member.