Online Learning for Offloading and Autoscaling in Energy Harvesting Mobile Edge Computing
Open accessThe DOI matches its Crossref record; 34 of the 34 references it cites were found in a registry or an index.
Abstract
Mobile edge computing (also known as fog computing) has recently emerged to enablein-situprocessing of delay-sensitive applications at the edge of mobile networks. Providing grid power supply in support of mobile edge computing, however, is costly and even infeasible (in certain rugged or under-developed areas), thus mandating on-site renewable energy as a major or even sole power supply in increasingly many scenarios. Nonetheless, the high intermittency and unpredictability of renewable energy make it very challenging to deliver a high quality of service to users in energy harvesting mobile edge computing systems. In this paper, we address the challenge of incorporating renewables into mobile edge computing and propose an efficient reinforcement learning-based resource management algorithm, which learns on-the-fly the optimal policy of dynamic workload offloading (to the centralized cloud) and edge server provisioning to minimize the long-term system cost (including both service delay and operational cost). Our online learning algorithm uses a decomposition of the (offline) value iteration and (online) reinforcement learning, thus achieving a significant improvement of learning rate and run-time performance when compared to standard reinforcement learning algorithms such as${Q}$-learning. We prove the convergence of the proposed algorithm and analytically show that the learned policy has a simple monotone structure amenable to practical implementation. Our simulation results validate the efficacy of our algorithm, which significantly improves the edge computing performance compared to fixed or myopic optimization schemes and conventional reinforcement learning algorithms.
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
“At this background, large state and action spaces in the whole offloading problem make the conventional learning algorithms [245] [256] [247] infeasible actually.”
From Convergence of Edge Computing and Deep Learning: A Comprehensive Survey (2020) · cites this paper as [245]
Citations
Cites
Reinforcement Learning: An Introduction
Sutton, R. S., & Barto, A. G. (1998) · IEEE Transactions on Neural Networks · doi:10.1109/tnn.1998.712192 · Cited by 27,225 · cited 5×
“The proposed PDS-based learning algorithm is compared with three benchmark schemes: • Q-learning [10]: Q-learning is a famous model-free reinforcement learning technique for solving MDP problems.”
BackgroundMethodInfluentialReinforcement Learning: An Introduction
(2005) · IEEE Transactions on Neural Networks · doi:10.1109/tnn.2004.842673 · Cited by 24,496
(1995) · Cited by 10,558
Edge computing: Vision and challenges
Shi, W., Cao, J., Zhang, Q., Li, Y., & Xu, L. (2016) · IEEE Internet of Things Journal · doi:10.1109/jiot.2016.2579198 · Cited by 8,733 · cited 4×
“backbone networks [1] but also result in (sometimes intolerable) large transmission latencies that degrade the quality of service [2]–[4].”
BackgroundInfluentialMarkov decision processes: Discrete stochastic dynamic programming
Hazeghi, K., & Puterman, M. L. (1995) · Journal of the American Statistical Association · doi:10.2307/2291177 · Cited by 8,037 · cited 1×
“battery state b and the power demand action a for any given λ,e,h. This is to prove that ∀b ≤ b′ and a ≤ a′, we have C(b,a ′)− C(b,a) ≥ C(b ,a′)− C(b′,a) (32) If the above is true, then we can apply [27] (Section 4.7) to show that π(s) is non-decreasing in b. Equation (32) is equivalent to V∗(b− a ′)− V ∗(b −a) ≥ V∗(b…”
InfluentialA 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
Markov Decision Processes: Discrete Stochastic Dynamic Programming
Baxter, L. A. (1995) · Technometrics · doi:10.1080/00401706.1995.10484354 · Cited by 2,415
Fog and IoT: An Overview of Research Opportunities
Chiang, M., & Zhang, T. (2016) · IEEE Internet of Things Journal · doi:10.1109/jiot.2016.2584538 · Cited by 2,362 · cited 2×
“Fog computing [5] is a related concept that refers to the same computing paradigm.”
BackgroundStochastic Network Optimization with Application to Communication and Queueing Systems
Neely, M. J. (2010) · Synthesis lectures on communication networks · doi:10.2200/s00271ed1v01y201006cnt007 · Cited by 2,218
Energy Harvesting Sensor Nodes: Survey and Implications
Sudevalayam, S., & Kulkarni, P. (2011) · IEEE Communications Surveys & Tutorials · doi:10.1109/surv.2011.060710.00094 · Cited by 1,981 · cited 2×
“and/or solar radiation is embraced as a major or even sole power supply for edge systems in the field, thanks to the recent advancements of energy harvesting techniques [8], [9].”
BackgroundDynamic computation offloading for mobile-edge computing with energy harvesting devices
Mao, Y., Zhang, J., & Letaief, K. B. (2016) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2016.2611964 · Cited by 1,705 · cited 1×
“For energy harvesting mobile devices, a dynamic computation offloading policy was proposed in [20] using Lyapunov optimization tech-”
MethodFinding your Way in the Fog: Towards a Comprehensive Definition of Fog Computing
Vaquero, L. M., & Rodero-Merino, L. (2014) · ACM SIGCOMM Computer Communication Review · doi:10.1145/2677046.2677052 · Cited by 1,066
Energy Harvesting Wireless Communications: A Review of Recent Advances
Ulukus, S., Yener, A., Erkip, E., Simeone, O., Zorzi, M., Grover, P., & Huang, K. (2015) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2015.2391531 · Cited by 805 · cited 2×
“and/or solar radiation is embraced as a major or even sole power supply for edge systems in the field, thanks to the recent advancements of energy harvesting techniques [8], [9].”
BackgroundEnergy Harvesting Wireless Communications: A Review of Recent Advances
Ulukuş, Ş., Yener, A., Erkip, E., Simeone, O., Member, S., Zorzi, M., Grover, P., Huang, K., & Member, S. (2015) · Cited by 643
Toward dynamic energy-efficient operation of cellular network infrastructure
Oh, E., Krishnamachari, B., Liu, X., & Niu, Z. (2011) · IEEE Communications Magazine · doi:10.1109/mcom.2011.5783985 · Cited by 562 · cited 1×
“However, designing green MEC powered by renewable energy is much more challenging compared to green communication systems [13]–[15] or green data center networks [16] [7], [17], [18] since the radio and computation resources have to be jointly managed, whereas prior research typically only considers one of the two dec…”
BackgroundThe O.D.E. Method for Convergence of Stochastic Approximation and Reinforcement Learning
Borkar, V. S., & Meyn, S. P. (2000) · SIAM Journal on Control and Optimization · doi:10.1137/s0363012997331639 · Cited by 499 · cited 1×
“It is proven in [26] that the convergence of our proposed algorithm is equivalent to the convergence of the associated ordinary differential equation:”
Dynamic right-sizing for power-proportional data centers
Lin, M., Wierman, A., Andrew, L. L. H., & Thereska, E. (2011) · 2011 Proceedings IEEE INFOCOM · doi:10.1109/infcom.2011.5934885 · Cited by 437
Dynamic Base Station Switching-On/Off Strategies for Green Cellular Networks
Oh, E., Son, K., & Krishnamachari, B. (2013) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2013.032013.120494 · Cited by 418 · cited 1×
“x λ(x,t)/θ(x,t), which results in a total wireless access and transmission delay of cwi(t) = P x λ(x,t)/[θ(x,t)(1−ρ(t))] by following the literature and modeling the base station as a queueing system [14]. Next we model the workload processing delay incurred at the edge servers. For the local processed workload, the d…”
MethodGreenSlot: scheduling energy consumption in green datacenters
Goiri, Í., Le, K., Haque, M. E., Beauchea, R., Nguyen, T. D., Guitart, J., Torres, J., & Bianchini, R. (2011) · Proceedings of 2011 International Conference for High Performance Computing, Networking, Storage and Analysis · doi:10.1145/2063384.2063411 · Cited by 309 · cited 1×
“However, designing green MEC powered by renewable energy is much more challenging compared to green communication systems [13]–[16] or green data center networks [7], [17]–[19]”
BackgroundManaging cost, performance, and reliability tradeoffs for energy-aware server provisioning
Guenter, B., Jain, N., & Williams, C. (2011) · 2011 Proceedings IEEE INFOCOM · doi:10.1109/infcom.2011.5934917 · Cited by 235 · cited 1×
“[23], [24].”
Towards power consumption-delay tradeoff by workload allocation in cloud-fog computing
Deng, R., Lu, R., Lai, C., & Luan, T. H. (2015) · 2015 IEEE International Conference on Communications (ICC) · doi:10.1109/icc.2015.7248934 · Cited by 197 · cited 1×
“ation of the wireless channel and energy. The focus of the present paper is on energy harvesting MEC systems and our solution is based on reinforcement learning. Another study relevant to our work is [21], which also studies workload allocation/offloading in a cloud-fog computing system. However, unlike our energy harv…”
Mobile Edge Computing: Survey and Research Outlook.Preprint
Mao, Y., You, C., Zhang, J., Huang, K., & Letaief, K. B. (2017) · arXiv (Cornell University) · Cited by 173
Han, T., & Ansari, N. (2015) · IEEE/ACM Transactions on Networking · doi:10.1109/tnet.2015.2404576 · Cited by 148 · cited 2×
“For instance, in many developing countries, the majority of base stations have to be powered by continuously operating diesel generators because the electric grid is too unreliable [7].”
BackgroundLi, C., Qouneh, A., & Li, T. (2012) · ACM SIGARCH Computer Architecture News · doi:10.1145/2366231.2337218 · Cited by 115 · cited 1×
“…MEC powered by renewable energy is much more challenging compared to green communication systems [13]–[15] or green data center networks [16] [7], [17], [18] since the radio and computation resources have to be jointly managed, whereas prior research typically only considers one of the two…”
BackgroundOnline Learning for Offloading and Autoscaling in Renewable-Powered Mobile Edge Computing
Xu, J., & Ren, S. (2016) · 2016 IEEE Global Communications Conference (GLOBECOM) · doi:10.1109/glocom.2016.7842069 · Cited by 90 · cited 2×
“To our best knowledge, the conference version of this paper [23] was the first to study resource management for energy harvesting MEC servers (see related discussions in a recent comprehensive survey paper [12]).”
BackgroundResultSelf-citationSocially trusted collaborative edge computing in ultra dense networks
Chen, L., & Xu, J. (2017) · Proceedings of the Second ACM/IEEE Symposium on Edge Computing · doi:10.1145/3132211.3134451 · Cited by 78
Self-citationMobile Edge Computing: Challenges for Future Virtual Network Embedding Algorithms
Beck, M. T., & Maier, M. (2014) · Cited by 55
Towards sustainable in-situ server systems in the big data era
Li, C., Hu, Y., Liu, L., Gu, J., Song, M., Liang, X., Yuan, J., & Li, T. (2015) · Proceedings of the 42nd Annual International Symposium on Computer Architecture · doi:10.1145/2749469.2750381 · Cited by 54 · cited 3×
“bines can combine their output to power the edge system and charge the batteries [6].”
BackgroundInfluentialOnline Geographical Load Balancing for Energy-Harvesting Mobile Edge Computing
Wu, H., Chen, L., Shen, C., Wen, W., & Xu, J. (2018) · 2018 IEEE International Conference on Communications (ICC) · doi:10.1109/icc.2018.8422299 · Cited by 49
Self-citationHoang, D. T., Niyato, D. T., Wang, P., & Kim, D. I. (2015) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2015.2488620 · Cited by 38
Structure-Aware Stochastic Storage Management in Smart Grids
Zhang, Y., & van der Schaar, M. (2014) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2014.2346477 · Cited by 30 · cited 2×
“The proof follows [24].”
Towards sustainable in-situ server systems in the big data era
Li, C., Hu, Y., Liu, L., Gu, J., Song, M., Liang, X., Yuan, J., & Li, T. (2015) · ACM SIGARCH Computer Architecture News · doi:10.1145/2872887.2750381 · Cited by 24
Cooperative Virtual Machine Management in Smart Grid Environment
Kaewpuang, R., Chaisiri, S., Niyato, D. T., Lee, B., & Wang, P. (2013) · IEEE Transactions on Services Computing · doi:10.1109/tsc.2013.37 · Cited by 21 · cited 1×
“However, designing green MEC powered by renewable energy is much more challenging compared to green communication systems [13]–[16] or green data center networks [7], [17]–[19]”
BackgroundOnline Geographical Load Balancing for Energy-Harvesting Mobile Edge ComputingPreprint
Xu, J., Wu, H., Chen, L., & Shen, C. (2017) · arXiv (Cornell University) · Cited by 2
Self-citation
Research works printed but not linked: 4
- reiss, 2011, Google Cluster-usage Traces Format+ Schema, 1
- patel, 2014, Mobile-Edge Computing—Introductory Technical White Paper
- 2017, Elia Power Generation
- rivera, 2014, Gartner Says the Internet of Things Will Transform the Data Center
Cited by
Ensemble deep learning: A review
Ganaie, M. A., Hu, M., Malik, A. K., Tanveer, M., & Suganthan, P. N. (2022) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2022.105151 · Cited by 2,143
“With the limitation of getting the complete data in real-world problems, online/incremental learning has been applied to various tasks like learning social representations [174], fog computing [175], identifying suspicious URLs [176], etc.”
MethodConvergence of Edge Computing and Deep Learning: A Comprehensive Survey
Wang, X., Han, Y., Leung, V. C. M., Niyato, D., Yan, X., & Chen, X. (2020) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2020.2970550 · Cited by 1,498
“At this background, large state and action spaces in the whole offloading problem make the conventional learning algorithms [245], [256], [247] infeasible actually.”
BackgroundPham, Q.-V., Fang, F., Ha, V. N., Piran, M. J., Le, M., Le, L. B., Hwang, W.-J., & Ding, Z. (2020) · IEEE Access · doi:10.1109/access.2020.3001277 · Cited by 870
“The literature [106] jointly studied the offloading and autoscaling policy (i.”
Computation Rate Maximization in UAV-Enabled Wireless-Powered Mobile-Edge Computing Systems
Zhou, F., Wu, Y., Hu, R. Q., & Qian, Y. (2018) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2018.2864426 · Cited by 851
Chen, X., Zhang, H., Wu, C., Mao, S., Ji, Y., & Bennis, M. (2019) · IEEE Internet of Things Journal · doi:10.1109/jiot.2018.2876279 · Cited by 680
Metadata
- Authors
- Xu, J., Chen, L., & Ren, S.
- Affiliations
- University of Miami (US)
- University of California, Riverside (US)
- Venue
- IEEE Transactions on Cognitive Communications and Networking
- Published
- 2017-09
- Volume
- 3
- Issue
- 3
- Pages
- 361-373
- ISSN
- 2332-7731 (ISSN-L)
- OpenAlex
- W2599106590
- Cited by
- 368
- References
- 33
- Citations by year
- 2026: 14 · 2025: 15 · 2024: 32 · 2023: 24 · 2022: 41 · 2021: 55 · 2020: 87 · 2019: 59 · 2018: 34 · 2017: 7
- FWCI
- 34.02
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- IEEE Transactions on Cognitive Communications and Networking (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- U.S. NSF (CNS-1551661, CNS-1565474, ECCS-1610471)
- Topics
- Physical Sciences › Computer Science › Computer Networks and Communications › IoT and Edge/Fog Computing
- Physical Sciences › Computer Science › Hardware and Architecture › Energy Efficiency in Computing
- Physical Sciences › Computer Science › Information Systems › Cloud Computing and Resource Management
- Keywords
- mobile edge computing, fog computing, energy harvesting, renewable energy, workload offloading, auto-scaling, reinforcement learning, online learning, service delay, operating costs, convergence analysis
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.