Deep Learning for Edge Computing Applications: A State-of-the-Art Survey
Open accessThe DOI matches its Crossref record; 137 of the 139 references it cites were found in a registry or an index; 2 of the 39 sources its claims cite were checked for support.
Abstract
With the booming development of Internet-of-Things (IoT) and communication technologies such as 5G, our future world is envisioned as an interconnected entity where billions of devices will provide uninterrupted service to our daily lives and the industry. Meanwhile, these devices will generate massive amounts of valuable data at the network edge, calling for not only instant data processing but also intelligent data analysis in order to fully unleash the potential of the edge big data. Both the traditional cloud computing and on-device computing cannot sufficiently address this problem due to the high latency and the limited computation capacity, respectively. Fortunately, the emerging edge computing sheds a light on the issue by pushing the data processing from the remote network core to the local network edge, remarkably reducing the latency and improving the efficiency. Besides, the recent breakthroughs in deep learning have greatly facilitated the data processing capacity, enabling a thrilling development of novel applications, such as video surveillance and autonomous driving. The convergence of edge computing and deep learning is believed to bring new possibilities to both interdisciplinary researches and industrial applications. In this article, we provide a comprehensive survey of the latest efforts on the deep-learning-enabled edge computing applications and particularly offer insights on how to leverage the deep learning advances to facilitate edge applications from four domains, i.e., smart multimedia, smart transportation, smart city, and smart industry. We also highlight the key research challenges and promising research directions therein. We believe this survey will inspire more researches and contributions in this promising field.
Overview
CoChat Summary
TL;DR
This survey reviews deep-learning-enabled edge computing applications across smart multimedia, smart transportation, smart city, and smart industry. It argues edge computing reduces latency and improves privacy/security while deep learning enables intelligent edge applications, and it highlights model training/inference, application enhancement, and hardware/software optimization as key challenges.
Key findings
- Edge computing advantages include low latency, energy saving, context-aware service, and privacy/security; the paper states cloud transmission latency is usually tens or hundreds of milliseconds, while edge latency is usually several milliseconds or even microsecond level.
- Deep learning models covered include RBM/DBN, autoencoder, DNN, CNN, RNN/LSTM, and DRL; applications span video analytics, adaptive streaming, caching, autonomous driving, traffic prediction/control, smart home/building/grid, and smart manufacturing/industrial analysis.
- The paper cites Cisco’s estimate of more than 50 billion IoT devices connected by 2020, and states global video traffic accounted for 75% of Internet traffic in 2017 and is estimated to grow four-fold by 2022.
- SignFi is reported to identify 276 sign language gestures including head, arm, hand, and finger using WiFi CSI signals with CNN classification.
Methods
- Design: a state-of-the-art survey/literature review, not an empirical study; it organizes deep-learning-enabled edge computing applications into four domains: smart multimedia, smart transportation, smart city, and smart industry.
- Data or sample: reviews prior research works and representative systems; no primary dataset or experimental sample is used.
- Main analysis: narrative synthesis of representative deep learning models and edge applications, with summary tables (Tables 1–4) and discussion of research challenges and directions.
Limitations
- The paper does not explicitly state limitations of the survey itself. It notes the field is still in the infant stage and that applications such as real-time VR gaming and autonomous driving still have a long way to go before practical application.
- Evident from the text: no systematic search protocol, inclusion/exclusion criteria, or quantitative meta-analysis are described; coverage is a narrative selection of representative works.
- Evident from the text: no primary experiments or performance benchmarks are conducted; reported numbers are cited from other sources.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures
Verification
Claims the paper makes
2 Valid · 6 Not valid · 25 Not checked · 38 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“And deep learning further empowers more effective local analysis and prediction [130] at the edge node of industry instead of the cloud.”
§ Deep Learning for Edge Computing Applications: A State-of-the-Art Survey > E. SMART INDUSTRY
What the sources say
- Partly supportedcites [130] matched to “CS2P: Improving Video Bitrate Selection and Adaptation with Data-Driven Throughput Prediction” (2016) sourceExperimental — accuracy not yet measured
Our copy supports only the 'local prediction instead of centralized/cloud control' half of the sentence: §4 (CS2P Approach and Overview) casts CS2P as 'a middle ground between centralized video control platforms (e.g., C3, CFA) and decentralized player-based bitrate adaptation' with 'actuation ... in a decentralized manner'; §5.3 (Player integration) describes each client downloading its compact HMM and predicting locally ('<10 milliseconds ... <5KB memory'), arguing 'the client is often in the best position to quickly detect performance issues'; §8 (Related Work) contrasts CFA's 'centralized control' with CS2P being 'amenable of a decentralized execution'. What the source does NOT support is the 'deep learning' attribution and the 'edge node of industry' setting: the predictor is explicitly 'a simple but effective HMM-based predictor' (§5.2) with clustering plus median-based initial prediction (§5.1, §5.2), and its learned-model baselines are classical machine learning only — SVR and GBR (§7.1); no neural network or deep-learning method appears anywhere in our copy, and the evaluated system is HTTP/DASH video clients and CDN/video servers, not industrial edge nodes. (LLM, conf=0.78)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
A statement the paper makes itself — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“With the advantage of the proximity nature, edge servers can collect more context information to support the data processing.”
§ Deep Learning for Edge Computing Applications: A State-of-the-Art Survey > 3) CONTEXT-AWARE SERVICE
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
In the paper
“Zhou et al. [13] mainly focused on the deep learning model training and inference with edge computing.”
§ Deep Learning for Edge Computing Applications: A State-of-the-Art Survey > I. INTRODUCTION
What the sources say
- Not checked yetcites [13] matched to “Restricted Boltzmann machines for collaborative filtering” (2007) source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
None of the sources cited for this claim have been checked yet.
In the paper
“There are also a series of surveys for mobile edge computing [2], [14]- [16] and deep learning [17], [18], respectively, while they focused on either of them without a comprehensive review on the combination.”
§ Deep Learning for Edge Computing Applications: A State-of-the-Art Survey > I. INTRODUCTION
What the sources say
- Not checked yetcites [2]
- Not checked yetcites [14]- [16]
- Not checked yetcites [17] matched to “Edge Computing for Autonomous Driving: Opportunities and Challenges” (2019) source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
In the paper
“Han et al. [11] presented their understanding on edge computing and deep learning from five aspects, while they did not make a comprehensive and in-depth overview from the perspective of applications.”
§ Deep Learning for Edge Computing Applications: A State-of-the-Art Survey > I. INTRODUCTION
What the sources say
- Not checked yetcites [11]
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
18 more claims
Create a free account to see all 71 claims and to verify the 37 references still open.
- Probably not a substantial claim. 38
- It may state two separate claims as one. 28
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?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“It is computationally expensive, consuming massive CPU and GPU resources, power, memory, and time.”
From TinyML: Enabling of Inference Deep Learning Models on Ultra-Low-Power IoT Edge Devices for AI Applications (2022) · cites this paper as [9]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“However, edge devices have not yet become cost-efficient enough to take training DL models due to limited resources [7,8,9].”
From TinyML: Enabling of Inference Deep Learning Models on Ultra-Low-Power IoT Edge Devices for AI Applications (2022) · cites this paper as [9]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“In addition, connecting IoT devices to the cloud to transfer raw data and perform processing causes delayed system responses, exposes private data and increases communication costs.”
From TinyML: Enabling of Inference Deep Learning Models on Ultra-Low-Power IoT Edge Devices for AI Applications (2022) · cites this paper as [9]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Deep Learning for Edge Computing Applications: A State-of-the-Art Survey > I. INTRODUCTION
“Han et al. [11] presented their understanding on edge computing and deep learning from five aspects”
§ Deep Learning for Edge Computing Applications: A State-of-the-Art Survey > References
“11. X. Wang, Y. Han, V. C. M. Leung, D. Niyato, X. Yan, and X. Chen, ''Convergence of edge computing and deep learning: A com- prehensive survey,''”
Citations
Cites
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318
LeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2577031 · Cited by 73,472
Learning representations by back-propagating errors
Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1986) · Nature · doi:10.1038/323533a0 · Cited by 31,892
Human-level control through deep reinforcement learning
Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A. A., Veness, J., Bellemare, M. G., Graves, A., Riedmiller, M., Fidjeland, A. K., Ostrovski, G., et al. (2015) · Nature · doi:10.1038/nature14236 · Cited by 30,357
YOLO9000: Better, Faster, Stronger
Redmon, J., & Farhadi, A. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.690 · Cited by 19,892
A fast learning algorithm for deep belief nets
Hinton, G. E., Osindero, S., & Teh, Y.-W. (2006) · Neural Computation · doi:10.1162/neco.2006.18.7.1527 · Cited by 16,336
Visualizing and understanding convolutional networks
Zeiler, M. D., & Fergus, R. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10590-1_53 · Cited by 15,839
Mastering the game of Go with deep neural networks and tree search
Silver, D., Huang, A., Maddison, C. J., Guez, A., Sifre, L., van den Driessche, G., Schrittwieser, J., Antonoglou, I., Panneershelvam, V., Lanctot, M., et al. (2016) · Nature · doi:10.1038/nature16961 · Cited by 15,612
Research works printed but not linked: 8
- I. Goodfellow, Y. Bengio, and A. Courville, Deep Learning. Cambridge, MA, USA: MIT Press, 2016. [Online]. Available: http://www. deeplearningbook.org
- Y. C. Hu, M. Patel, D. Sabella, N. Sprecher, and V. Young, ''Mobile edge computing-A key technology towards 5G,'' ETSI white paper, vol. 11, no. 11, pp. 1-16, 2015.
- F. Wang, F. Wang, X. Fan, and J. Liu, ''BatAlloc: Effective battery allocation against power outage for cellular base stations,'' in Proc. ACM e-Energy, 2017, pp. 234-241.
- X. Song, H. Kanasugi, and R. Shibasaki, ''DeepTransport: Prediction and simulation of human mobility and transportation mode at a citywide level,'' in Proc. IJCAI, vol. 16, 2016, pp. 2618-2624.
- Cisco Visual Networking Index: Forecast and Trends, 2017 to 2022 White Paper. Accessed: Mar. 21, 2020. [Online]. Available: https://www.cisco.com/c/en/us/solutions/collateral/service-provider/ visual-networking-index-vni/white-paper-c11-741490.html
- FANGXIN WANG (Student Member, IEEE) received the B.S. degree from the Department of Computer Science of Technology, Beijing Uni- versity of Post and Telecommunication, Beijing, China, in 2013, and the M.S. degree from the Department of Computer Science and Technol- ogy, Beijing, in 2016. He is currently pursuing the Ph.D. degree with the School of Comput- ing Science, Simon Fraser University, Burnaby, Canada. His research interests include the Internet of Things, wireless networking, multimedia systems, big data analysis, and machine learning.
- JIANGCHUAN LIU (Fellow, IEEE) received the B.Eng. (cum laude) degree in computer science from Tsinghua University, Beijing, China, in 1999, and the Ph.D. degree in computer science from The Hong Kong University of Science and Technology, in 2003. He is currently a Full Professor (with University Professorship) with the School of Computing Sci- ence, Simon Fraser University, BC, Canada. He is a Fellow of the Canadian Academy of Engineer- ing and the NSERC E.W.R. Steacie Memorial Fellow. He is a Steering Committee Member of IEEE TRANSACTIONS ON MOBILE COMPUTING. He was a co-recipient of the Test of Time Paper Award of the IEEE INFOCOM, in 2015, the ACM TOMCCAP Nicolas D. Georganas Best Paper Award, in 2013, and the ACM Multimedia Best Paper Award, in 2012. He is an Associate Editor of the IEEE/ACM TRANSACTIONS ON NETWORKING, the IEEE TRANSACTIONS ON BIG DATA, and the IEEE TRANSACTIONS ON MULTIMEDIA.
- T. Huang, R.-X. Zhang, C. Zhou, and L. Sun, ''QARC: Video quality aware rate control for real-time video streaming based on deep reinforce- ment learning,'' in Proc. ACM Multimedia, 2018, pp. 1208-1216.
Cited by
Jusup, M., Holme, P., Kanazawa, K., Takayasu, M., Romić, I., Wang, Z., Geček, S., Lipić, T., Podobnik, B., Wang, L., et al. (2022) · Physics Reports · doi:10.1016/j.physrep.2021.10.005 · Cited by 643
Federated learning review: Fundamentals, enabling technologies, and future applications
Banabilah, S., Aloqaily, M., Alsayed, E., Malik, N., & Jararweh, Y. (2022) · Information Processing & Management · doi:10.1016/j.ipm.2022.103061 · Cited by 556
Alajlan, N. N., & Ibrahim, D. M. (2022) · Micromachines · doi:10.3390/mi13060851 · Cited by 224
“Currently, DL is partially deployed on edge devices and the remaining data are transferred to processing in the cloud, or DL models are deployed in the cloud to process raw data that are received from edge devices, thus causing a delay in latency [9].”
MethodEdge Computing for IoT-Enabled Smart Grid: The Future of Energy
Minh, Q. N., Nguyen, V.-H., Quy, V. K., Ngoc, L. A., Chehri, A., & Jeon, G. (2022) · Energies · doi:10.3390/en15176140 · Cited by 170
BUSINESS INTELLIGENCE IN THE ERA OF BIG DATA: A REVIEW OF ANALYTICAL TOOLS AND COMPETITIVE ADVANTAGE
Adewusi, A. O., Okoli, U. I., Adaga, E. M., Olorunsogo, T., Asuzu, O. F., & Daraojimba, D. O. (2024) · Computer Science & IT Research Journal · doi:10.51594/csitrj.v5i2.791 · Cited by 200
Metadata
- Authors
- Wang, F., Zhang, M., Wang, X., Ma, X., & Liu, J.
- Affiliations
- Simon Fraser University (CA)
- Huazhong University of Science and Technology (CN)
- Venue
- IEEE Access
- Published
- 2020
- Volume
- 8
- Pages
- 58322-58336
- ISSN
- 2169-3536 (ISSN-L)
- OpenAlex
- W3013557734
- Semantic Scholar
- a4513c402c12fd495dce583522d98f9a695e0994
- Cited by
- 259
- References
- 139
- Citations by year
- 2026: 37 · 2025: 67 · 2024: 55 · 2023: 27 · 2022: 45 · 2021: 23 · 2020: 5
- FWCI
- 11.75
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Locations
- IEEE Access (publishedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Summit (Simon Fraser University) (submittedVersion, cc-by)
- Funders
- Canada NSERC Discovery Grant and
- National Natural Science Foundation of China (61702204)
- Topics
- Physical Sciences › Computer Science › Computer Networks and Communications › IoT and Edge/Fog Computing
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Information Systems › Internet of Things and AI
- Keywords
- edge computing, deep learning, Internet of Things, 5G, latency reduction, video surveillance, autonomous driving, smart transportation, smart cities, smart industry
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.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-07, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-07 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (qwen3.8-omni-flash).
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Deep Learning for Edge Computing Applications: A State-of-the-Art Survey. https://app.cochat.ai/papers/10.1109/access.2020.2982411
- The paper
- Wang, F., Zhang, M., Wang, X., Ma, X., & Liu, J. (2020). Deep Learning for Edge Computing Applications: A State-of-the-Art Survey. IEEE Access. https://doi.org/10.1109/access.2020.2982411
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.






