Distributed Multi-Cloud Multi-Access Edge Computing by Multi-Agent Reinforcement Learning
The DOI matches its Crossref record; 44 of the 44 references it cites were found in a registry or an index.
Abstract
In this paper, we consider a three-layer distributed multi-access edge computing (MEC) network where multiple clouds, MEC servers, and edge devices (EDs) are deployed at the top layer, middle layer, and bottom layer, respectively. Each cloud center (CC) is associated with an independent service provider and publishes an application-driven computing task. To deliver the tasks, CCs rely on EDs to generate the raw data and offload part of the computing tasks to both EDs and MEC servers such that their computing and transmission resources can be fully utilized to reduce the system latency. However, in such a three-layer network, the distributed deployment of tasks leads to inevitable resource competition among CCs. To address this issue, we propose a distributed scheme based on multi-agent reinforcement learning, where each CC jointly determines the task offloading and resource allocation strategy based on its inference of other CCs’ decisions. Simulation results indicate that a lower system latency is achieved via our proposed scheme compared with the existing schemes. In addition, the influence of the number of CCs, MEC servers, and EDs on latency performance is also discussed.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Besides, a multi-agent RL approach was developed in [260] to solve the distributed mixed combinatorial optimization problem for task offloading and resource allocation in multi-layer edge inference systems.”
From Edge artificial intelligence for 6g: vision, enabling technologies, and applications (2022) · cites this paper as [260]
Citations
Cites
Fundamentals of Wireless Communication
Tse, D., & Viswanath, P. (2005) · doi:10.1017/cbo9780511807213 · Cited by 13,162 · cited 1×
“Note that the orthogonal resources, e.g., spectrum and time resources, are considered as the wireless transmission resources in this paper which influence the wireless data rate in a linear manner [34].”
BackgroundArmbrust, M., Fox, A., Griffith, R., Joseph, A. D., Katz, R., Konwinski, A., Lee, G., Patterson, D., Rabkin, A., Stoica, I., et al. (2010) · Communications of the ACM · doi:10.1145/1721654.1721672 · Cited by 9,464 · cited 1×
“These applications are generally processed in the remote cloud centers (CCs), such as Amazon EC2, which is called cloud computing [4]–[6].”
BackgroundWatkins, C. J. C. H., & Dayan, P. (1992) · Machine Learning · doi:10.1007/bf00992698 · Cited by 9,083
Reinforcement Learning: A Survey
Kaelbling, L. P., Littman, M. L., & Moore, A. W. (1996) · Journal of Artificial Intelligence Research · doi:10.1613/jair.301 · Cited by 8,972 · cited 1×
“…task offloading [27], [28] scheme for MEC systems under a stochastic and dynamic environment, several initial works have considered reinforcement learning (RL) as a promising technique since it takes the future feedback from the environment into account, thereby achieving a long-term goal [29].”
BackgroundA Survey on Mobile Edge Computing: The Communication Perspective
Mao, Y., You, C., Zhang, J., Huang, K., & Letaief, K. B. (2017) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2017.2745201 · Cited by 5,607 · cited 1×
“…with the issue and support more access modes including wired and wireless access in heterogeneous networks [11], [12], the concept of multi-access edge computing (MEC) has been proposed to enable the computing tasks to be offloaded to MEC servers which are deployed between the EDs and CCs [13]–[15].”
BackgroundMobile Edge Computing: A Survey on Architecture and Computation Offloading
Mach, P., & Becvar, Z. (2017) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2017.2682318 · Cited by 3,258 · cited 1×
“To cope with the issue and support more access modes including wired and wireless access in heterogeneous networks [11], [12], the concept of multi-access edge computing (MEC) has been proposed to enable the computing tasks to be offloaded to MEC servers which are deployed between the EDs and CCs [13]–[15].”
BackgroundEfficient multi-user computation offloading for mobile-edge cloud computing
Chen, X., Jiao, L., Li, W., & Fu, X. (2016) · IEEE/ACM Transactions on Networking · doi:10.1109/tnet.2015.2487344 · Cited by 2,675 · cited 2×
“In the literature, existing works have studied the task offloading problem either in a two-layer ED-MEC network [16]–[22], or in the three-layer ED-MEC-CC network with a single CC [23]–[25].”
BackgroundMethodA survey of mobile cloud computing: architecture, applications, and approaches
Dinh, H. T., Lee, C., Niyato, D., & Wang, P. (2013) · Wireless Communications and Mobile Computing · doi:10.1002/wcm.1203 · Cited by 2,391 · cited 1×
“These applications are generally processed in the remote cloud centers (CCs), such as Amazon EC2, which is called cloud computing [4]–[6].”
BackgroundTaleb, T., Samdanis, K., Mada, B., Flinck, H., Dutta, S., & Sabella, D. (2017) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2017.2705720 · Cited by 1,775 · cited 1×
“To cope with the issue and support more access modes including wired and wireless access in heterogeneous networks [11], [12], the concept of multi-access edge computing (MEC) has been proposed to enable the computing tasks to be offloaded to MEC servers which are deployed between the EDs and CCs…”
BackgroundThe dynamics of reinforcement learning in cooperative multiagent systems
Claus, C., & Boutilier, C. E. (1998) · National Conference on Artificial Intelligence · Cited by 1,463
Joint Task Offloading and Resource Allocation for Multi-Server Mobile-Edge Computing Networks
Tran, T. X., & Pompili, D. (2019) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2018.2881191 · Cited by 1,046
Mobile-Edge Computing: Partial Computation Offloading Using Dynamic Voltage Scaling
Wang, Y., Sheng, M., Wang, X., Wang, L., & Li, J. (2016) · IEEE Transactions on Communications · doi:10.1109/tcomm.2016.2599530 · Cited by 1,002 · cited 3×
“In the literature, existing works have studied the task offloading problem either in a two-layer ED-MEC network [16]–[22], or in the three-layer ED-MEC-CC network with a single CC [23]–[25].”
BackgroundBi, S., & Zhang, Y. J. (2018) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2018.2821664 · Cited by 870
Self-citationOffloading in Mobile Edge Computing: Task Allocation and Computational Frequency Scaling
Thinh, T. Q., Tang, J., La, Q. D., & Quek, T. Q. S. (2017) · IEEE Transactions on Communications · doi:10.1109/tcomm.2017.2699660 · Cited by 866 · cited 2×
“In the former, authors in [16]–[19] studied task offloading from EDs to a single MEC server, while MEC networks with multiple MEC servers were considered in [20]–[22].”
BackgroundZhang, K., Mao, Y., Leng, S., He, Y., & ZHANG, Y. (2017) · IEEE Vehicular Technology Magazine · doi:10.1109/mvt.2017.2668838 · Cited by 701 · cited 1×
“For example, in vehicle networks, EDs collect data about road information and upload to a remote CC, thereby constructing a detailed map of the real-time traffic conditions on road congestion and lane closures ahead [7].”
BackgroundDi, B., Zhang, H., Song, L., Li, Y., Han, Z., & Poor, H. V. (2020) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2020.3000813 · Cited by 544
Self-citationLatency Optimization for Resource Allocation in Mobile-Edge Computation Offloading
Ren, J., Yu, G., Cai, Y., & He, Y. (2018) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2018.2845360 · Cited by 450
Bringing IoT and Cloud Computing towards Pervasive Healthcare
Doukas, C., & Maglogiannis, I. (2012) · 2012 Sixth International Conference on Innovative Mobile and Internet Services in Ubiquitous Computing · doi:10.1109/imis.2012.26 · Cited by 413
Online Learning for Offloading and Autoscaling in Energy Harvesting Mobile Edge Computing
Xu, J., Chen, L., & Ren, S. (2017) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2017.2725277 · Cited by 368 · cited 1×
“Note that the time interval between the publication of two tasks is much larger than the duration for obtaining the optimal scheme [37].”
BackgroundWang, C., Yu, F. R., Liang, C., Chen, Q., & Tang, L. (2017) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2017.2672701 · Cited by 368
UAV-Enhanced Intelligent Offloading for Internet of Things at the Edge
Guo, H., & Liu, J. (2020) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2019.2954944 · Cited by 306 · cited 1×
“Moreover, to design vehicle charging [26], task offloading [27], [28] scheme for MEC”
MethodUltra-Dense LEO: Integrating Terrestrial-Satellite Networks Into 5G and Beyond for Data Offloading
Di, B., Zhang, H., Song, L., Li, Y., & Li, G. Y. (2019) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2018.2875980 · Cited by 286
Self-citationCloud Computing Security: From Single to Multi-clouds
AlZain, M. A., Pardede, E., Soh, B., & Thom, J. A. (2012) · 2012 45th Hawaii International Conference on System Sciences · doi:10.1109/hicss.2012.153 · Cited by 285 · cited 1×
“See https://www.ieee.org/publications/rights/index.html for more information. tend to launch their applications on different CCs [30]–[32] instead of just a single CC, 1 it is necessary to investigate the multi-CC case where each CC is associated with an independent service provider.”
BackgroundComputation Offloading for Multi-Access Mobile Edge Computing in Ultra-Dense Networks
Guo, H., Liu, J., & Zhang, J. (2018) · IEEE Communications Magazine · doi:10.1109/mcom.2018.1701069 · Cited by 183 · cited 1×
“To cope with the issue and support more access modes including wired and wireless access in heterogeneous networks [11], [12], the concept of multi-access edge computing (MEC) has been proposed to enable the computing tasks to be offloaded to MEC servers which are deployed between the EDs and CCs [13]–[15].”
BackgroundSmart and Resilient EV Charging in SDN-Enhanced Vehicular Edge Computing Networks
Liu, J., Guo, H., Xiong, J., Kato, N., Zhang, J., & Zhang, Y. (2020) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2019.2951966 · Cited by 177 · cited 1×
“Moreover, to design vehicle charging [26], task offloading [27], [28] scheme for MEC systems under a stochastic and dynamic environment, several initial works have considered reinforcement learning (RL) as a promising technique since it takes the future feedback from the environment into account,…”
BackgroundSelf-citationWireless Networks for Mobile Edge Computing: Spatial Modeling and Latency Analysis
Ko, S.-W., Han, K., & Huang, K. (2018) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2018.2840120 · Cited by 173
Wang, P., Yao, C., Zheng, Z., Sun, G., & Song, L. (2019) · IEEE Internet of Things Journal · doi:10.1109/jiot.2018.2876198 · Cited by 168 · cited 2×
“Authors in [25] combined the cloud and edge to construct a 3-layer data flow processing system consisting of multiple EDs, APs, and a single CC, such that the computing and transmission resources of the entire network can be fully utilized.”
BackgroundSelf-citationIntelligent Task Offloading in Vehicular Edge Computing Networks
Guo, H., Liu, J., Ren, J., & Zhang, Y. (2020) · IEEE Wireless Communications · doi:10.1109/mwc.001.1900489 · Cited by 155 · cited 1×
“Moreover, to design vehicle charging [26], task offloading [27], [28] scheme for MEC”
MethodSelf-citationMultiagent Reinforcement Learning for Multi-Robot Systems: A Survey
Yang, E., & Gu, D. (2004) · Cited by 139
Yang, L., Zhang, H., Li, X., Ji, H., & Leung, V. C. M. (2018) · IEEE/ACM Transactions on Networking · doi:10.1109/tnet.2018.2876941 · Cited by 137 · cited 2×
“In [24], a distributed computation offloading strategy was studied for a MEC system based on”
MethodCheng, K., Teng, Y., Sun, W., Liu, A., & Wang, X. (2018) · 2018 IEEE International Conference on Communications (ICC) · doi:10.1109/icc.2018.8422877 · Cited by 134
Classes of Multiagent Q-learning Dynamics with epsilon-greedy Exploration
Wunder, M. B., Littman, M. L., & Babes, M. (2010) · International Conference on Machine Learning · Cited by 117
Wang, P., Zheng, Z., Di, B., & Song, L. (2019) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2019.2931315 · Cited by 106
Self-citationEdge Computing and Social Internet of Things for Large-Scale Smart Environments Development
Cicirelli, F., Guerrieri, A., Spezzano, G., Vinci, A., Briante, O., Iera, A., & Ruggeri, G. (2018) · IEEE Internet of Things Journal · doi:10.1109/jiot.2017.2775739 · Cited by 88 · cited 1×
“W ITH the development of mobile communication networks, a number of latency-sensitive applications have emerged in the fields of transportation [1], healthcare [2], and smart environment [3], etc.”
BackgroundMulti-Cloud Provisioning and Load Distribution for Three-Tier Applications
Grozev, N., & Buyya, R. (2014) · ACM Transactions on Autonomous and Adaptive Systems · doi:10.1145/2662112 · Cited by 88
Dynamic analysis of multiagent Q -learning with ε-greedy exploration
Gomes, E. R., & Kowalczyk, R. (2009) · International Conference on Machine Learning · doi:10.1145/1553374.1553422 · Cited by 87 · cited 1×
“2) (cid:8) -Greedy Exploration Strategy: At the beginning of each iteration, each CC l selects an action a l according to the -greedy exploration strategy [40].”
MethodCooperative Collision Avoidance for Overtaking Maneuvers in Cellular V2X-Based Autonomous Driving
Deng, R., Di, B., & Song, L. (2019) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2019.2906509 · Cited by 81 · cited 1×
“W ITH the development of mobile communication networks, a number of latency-sensitive applications have emerged in the fields of transportation [1], healthcare [2], and smart environment [3], etc.”
BackgroundSelf-citationHetMEC: Heterogeneous Multi-Layer Mobile Edge Computing in the 6 G Era
Zhang, Y., Di, B., Wang, P., Lin, J., & Song, L. (2020) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2020.2975559 · Cited by 74 · cited 1×
“See https://www.ieee.org/publications/rights/index.html for more information. tend to launch their applications on different CCs [30]–[32] instead of just a single CC, 1 it is necessary to investigate the multi-CC case where each CC is associated with an independent service provider.”
BackgroundSelf-citationMachine learning for 5G and beyond: From model-based to data-driven mobile wireless networks
Wang, T., Wang, S., & Zhou, Z. (2019) · China Communications · doi:10.12676/j.cc.2019.01.015 · Cited by 70 · cited 1×
“However, an increasing number of mobile applications have posed great pressure on CCs and result in excessively high computing and transmission latency [8]–[10].”
BackgroundOpenRS-Cloud: A remote sensing image processing platform based on cloud computing environment
Guo, W., Gong, J., Jiang, W., Liu, Y., & She, B. (2010) · Science China Technological Sciences · doi:10.1007/s11431-010-3234-y · Cited by 54
Hilbert, M., López, P., & Vásquez, C. (2010) · The Information Society · doi:10.1080/01972241003712199 · Cited by 52 · cited 1×
“…paper, we simply use the computing rate to describe the computing resource, since it can be converted from any other kinds of computing resources [33], such as the number of virtual machines and CPU cycles per second. where x n,mED ( l ) denotes the percentage of the total raw data to be…”
MethodSadhu, A. K., & Konar, A. (2018) · IEEE Transactions on Systems Man and Cybernetics Systems · doi:10.1109/tsmc.2018.2865488 · Cited by 20 · cited 1×
“3) Complexity: The computational complexity is O(L|A|L) in each iteration [41], due to the calculation of the observed frequency distribution of other agents in (21) and (22).”
BackgroundLectures on learning and equilibrium in strategic form games
Drew Fudenberg, B., & Kreps, D. M. (1992) · Cited by 16
Sadhu, A. K., & Konar, A. (2020) · Wiley eBooks · doi:10.1002/9781119699057.ch4 · Cited by 7
Cited by
Edge artificial intelligence for 6g: vision, enabling technologies, and applications
Letaief, K. B., Shi, Y., Lu, J., & Lu, J. (2022) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2021.3126076 · Cited by 829
Alam, M., & Jamalipour, A. (2022) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2022.3160099 · Cited by 150
Blockchain-Empowered Collaborative Task Offloading for Cloud-Edge-Device Computing
Yao, S. F., Wang, M., Qu, Q., Zhang, Z., Zhang, Y.-F., Xu, K., & Xu, M. (2022) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2022.3213358 · Cited by 72
Self-citationLiu, Q., Zhang, H., Zhang, X., & Yuan, D. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3430486 · Cited by 61
“Most of works tackle this problem by optimizing the communication and computing resources to reduce delay and energy consumption [6], [7], [8], [9].”
BackgroundSuzuki, A., Kobayashi, M., & Oki, E. (2023) · IEEE Transactions on Network and Service Management · doi:10.1109/tnsm.2023.3267809 · Cited by 57
“Several studies have addressed task-offloading problems for CC and EC [1], [2], [3], [4], [5], [6], [7], [8].”
BackgroundMethodInfluential
Metadata
- Authors
- Zhang, Y., Di, B., Zheng, Z., Lin, J., & Song, L.
- Affiliations
- Peking University (CN)
- Imperial College London (GB)
- Venue
- IEEE Transactions on Wireless Communications
- Published
- 2021-04
- Volume
- 20
- Issue
- 4
- Pages
- 2565-2578
- ISSN
- 1536-1276 (ISSN-L)
- 1558-2248
- OpenAlex
- W3112164530
- Cited by
- 75
- References
- 44
- Citations by year
- 2026: 8 · 2025: 17 · 2024: 17 · 2023: 11 · 2022: 14 · 2021: 8
- FWCI
- 3.88
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- Locations
- IEEE Transactions on Wireless Communications (publishedVersion)
- Funders
- National Natural Science Foundation of China (61625101, 61829101, 61941101)
- NSF (EARS-1839818, CNS-1717454, CNS-1731424, CNS-1702850)
- Topics
- Physical Sciences › Computer Science › Computer Networks and Communications › IoT and Edge/Fog Computing
- Physical Sciences › Computer Science › Information Systems › Cloud Computing and Resource Management
- Physical Sciences › Computer Science › Hardware and Architecture › Energy Efficiency in Computing
- Keywords
- mobile edge computing, multi-cloud, task offloading, resource allocation, multi-agent reinforcement learning, three-layer network architecture, system latency, resource competition, decentralized decision-making, edge devices
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.