{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,1]],"date-time":"2026-04-01T10:50:33Z","timestamp":1775040633998,"version":"3.50.1"},"reference-count":50,"publisher":"MDPI AG","issue":"10","license":[{"start":{"date-parts":[[2019,9,29]],"date-time":"2019-09-29T00:00:00Z","timestamp":1569715200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100003593","name":"Conselho Nacional de Desenvolvimento Cient\u00edfico e Tecnol\u00f3gico","doi-asserted-by":"publisher","award":["142235\/2017-2"],"award-info":[{"award-number":["142235\/2017-2"]}],"id":[{"id":"10.13039\/501100003593","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100002322","name":"Coordena\u00e7\u00e3o de Aperfei\u00e7oamento de Pessoal de N\u00edvel Superior","doi-asserted-by":"publisher","award":["Finance Code 001"],"award-info":[{"award-number":["Finance Code 001"]}],"id":[{"id":"10.13039\/501100002322","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100002322","name":"Coordena\u00e7\u00e3o de Aperfei\u00e7oamento de Pessoal de N\u00edvel Superior","doi-asserted-by":"publisher","award":["PROCAD 2966\/2014"],"award-info":[{"award-number":["PROCAD 2966\/2014"]}],"id":[{"id":"10.13039\/501100002322","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001807","name":"Funda\u00e7\u00e3o de Amparo \u00e0 Pesquisa do Estado de S\u00e3o Paulo","doi-asserted-by":"publisher","award":["2013\/08293-7"],"award-info":[{"award-number":["2013\/08293-7"]}],"id":[{"id":"10.13039\/501100001807","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Future Internet"],"abstract":"<jats:p>Billions of devices will compose the IoT system in the next few years, generating a huge amount of data. We can use fog computing to process these data, considering that there is the possibility of overloading the network towards the cloud. In this context, deep learning can treat these data, but the memory requirements of deep neural networks may prevent them from executing on a single resource-constrained device. Furthermore, their computational requirements may yield an unfeasible execution time. In this work, we propose Deep Neural Networks Partitioning for Constrained IoT Devices, a new algorithm to partition neural networks for efficient distributed execution. Our algorithm can optimize the neural network inference rate or the number of communications among devices. Additionally, our algorithm accounts appropriately for the shared parameters and biases of Convolutional Neural Networks. We investigate the inference rate maximization for the LeNet model in constrained setups. We show that the partitionings offered by popular machine learning frameworks such as TensorFlow or by the general-purpose framework METIS may produce invalid partitionings for very constrained setups. The results show that our algorithm can partition LeNet for all the proposed setups, yielding up to 38% more inferences per second than METIS.<\/jats:p>","DOI":"10.3390\/fi11100209","type":"journal-article","created":{"date-parts":[[2019,9,30]],"date-time":"2019-09-30T05:58:33Z","timestamp":1569823113000},"page":"209","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":25,"title":["Partitioning Convolutional Neural Networks to Maximize the Inference Rate on Constrained IoT Devices"],"prefix":"10.3390","volume":"11","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8531-4559","authenticated-orcid":false,"given":"Fab\u00edola","family":"Martins Campos de Oliveira","sequence":"first","affiliation":[{"name":"Institute of Computing, University of Campinas, Campinas 13083-852, SP, Brazil"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1783-4231","authenticated-orcid":false,"given":"Edson","family":"Borin","sequence":"additional","affiliation":[{"name":"Institute of Computing, University of Campinas, Campinas 13083-852, SP, Brazil"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2019,9,29]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"27","DOI":"10.1145\/2677046.2677052","article-title":"Finding Your Way in the Fog: Towards a Comprehensive Definition of Fog Computing","volume":"44","author":"Vaquero","year":"2014","journal-title":"SIGCOMM Comput. Commun. Rev."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"16","DOI":"10.1109\/MCOM.2017.1600514","article-title":"Internet-of-Things-Based Smart Cities: Recent Advances and Challenges","volume":"55","author":"Mehmood","year":"2017","journal-title":"IEEE Commun. Mag."},{"key":"ref_3","unstructured":"Cisco Systems, I. (2019, July 22). Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update. Available online: https:\/\/www.cisco.com\/c\/en\/us\/solutions\/collateral\/service-provider\/visual-networking-index-vni\/white-paper-c11-738429.html."},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"Miraz, M.H., Ali, M., Excell, P.S., and Picking, R. (2018). Internet of Nano-Things, Things and Everything: Future Growth Trends. Future Internet, 10.","DOI":"10.3390\/fi10080068"},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"1125","DOI":"10.1109\/JIOT.2017.2683200","article-title":"A Survey on Internet of Things: Architecture, Enabling Technologies, Security and Privacy, and Applications","volume":"4","author":"Lin","year":"2017","journal-title":"IEEE Internet Things J."},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Bormann, C., Ersue, M., and Keranen, A. (2019, April 04). Terminology for Constrained-Node Networks. Available online: https:\/\/doi.org\/10.17487\/RFC7228.","DOI":"10.17487\/rfc7228"},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1186\/s40537-014-0007-7","article-title":"Deep learning applications and challenges in big data analytics","volume":"2","author":"Najafabadi","year":"2015","journal-title":"J. Big Data"},{"key":"ref_8","unstructured":"De Coninck, E., Verbelen, T., Vankeirsbilck, B., Bohez, S., Simoens, P., Demeester, P., and Dhoedt, B. (2015, January 26\u201327). Distributed neural networks for Internet of Things: The Big-Little approach. Proceedings of the 2nd EAI International Conference on Software Defined Wireless Networks and Cognitive Technologies for IoT, Rome, Italy."},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Grimaldi, M., Tenace, V., and Calimera, A. (2018). Layer-Wise Compressive Training for Convolutional Neural Networks. Future Internet, 11.","DOI":"10.3390\/fi11010007"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"355","DOI":"10.1016\/j.future.2019.03.001","article-title":"Multi-fidelity deep neural networks for adaptive inference in the internet of multimedia things","volume":"97","author":"Leroux","year":"2019","journal-title":"Future Gener. Comput. Syst."},{"key":"ref_11","unstructured":"Cortes, C., Lawrence, N.D., Lee, D.D., Sugiyama, M., and Garnett, R. (2015). Learning both Weights and Connections for Efficient Neural Network. Advances in Neural Information Processing Systems 28, Proceedings of the 29th Conference on Neural Information Processing Systems, Montr\u00e9al, QC, Canada, 7\u201312 December 2015, Curran Associates, Inc."},{"key":"ref_12","unstructured":"Guo, Y., Yao, A., and Chen, Y. (2016, January 5\u201310). Dynamic Network Surgery for Efficient DNNs. Proceedings of the 30th International Conference on Neural Information Processing Systems (NIPS\u201916), Barcelona, Spain."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Yao, S., Zhao, Y., Zhang, A., Su, L., and Abdelzaher, T. (2017, January 5\u20138). DeepIoT: Compressing Deep Neural Network Structures for Sensing Systems with a Compressor-Critic Framework. Proceedings of the 15th ACM Conference on Embedded Network Sensor Systems (SenSys \u201917), Delft, The Netherlands.","DOI":"10.1145\/3131672.3131675"},{"key":"ref_14","unstructured":"Abadi, M., Barham, P., Chen, J., Chen, Z., Davis, A., Dean, J., Devin, M., Ghemawat, S., Irving, G., and Isard, M. (2016, January 2\u20134). TensorFlow: A System for Large-scale Machine Learning. Proceedings of the 12th USENIX Conference on Operating Systems Design and Implementation (OSDI\u201916), Savannah, GA, USA."},{"key":"ref_15","unstructured":"De Coninck, E., Verbelen, T., Vankeirsbilck, B., Bohez, S., Leroux, S., and Simoens, P. (2015, January 7\u201311). DIANNE: Distributed Artificial Neural Networks for the Internet of Things. Proceedings of the 2nd Workshop on Middleware for Context-Aware Applications in the IoT (M4IoT 2015), Vancouver, BC, Canada."},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Lane, N.D., Bhattacharya, S., Georgiev, P., Forlivesi, C., Jiao, L., Qendro, L., and Kawsar, F. (2016, January 11\u201314). DeepX: A Software Accelerator for Low-Power Deep Learning Inference on Mobile Devices. Proceedings of the 15th ACM\/IEEE International Conference on Information Processing in Sensor Networks (IPSN), Vienna, Austria.","DOI":"10.1109\/IPSN.2016.7460664"},{"key":"ref_17","unstructured":"STMicroelectronics (2019, July 22). STM32 32-bit Arm Cortex MCUs. Available online: https:\/\/www.st.com\/en\/microcontrollers-microprocessors\/stm32-32-bit-arm-cortex-mcus.html."},{"key":"ref_18","unstructured":"Naumann, U., Schenk, O., Simon, H.D., and Toledo, S. (2009). Distillating knowledge about SCOTCH. Combinatorial Scientific Computing, Proceedings of the Dagstuhl Seminar, Dagstuhl, Germany, 3\u20138 May 2009, Schloss Dagstuhl-Leibniz-Zentrum fuer Informatik."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"359","DOI":"10.1137\/S1064827595287997","article-title":"A Fast and High Quality Multilevel Scheme for Partitioning Irregular Graphs","volume":"20","author":"Karypis","year":"1998","journal-title":"SIAM J. Sci. Comput."},{"key":"ref_20","unstructured":"De Oliveira, F.M.C., and Borin, E. (2018, January 24\u201327). Partitioning Convolutional Neural Networks for Inference on Constrained Internet-of-Things Devices. Proceedings of the 30th International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD), Lyon, France."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1016\/j.jnca.2017.12.001","article-title":"Distributed Data Stream Processing and Edge Computing","volume":"103","author":"Veith","year":"2018","journal-title":"J. Netw. Comput. Appl."},{"key":"ref_22","unstructured":"OpenFog Consortium Architecture Working Group (2019, July 22). OpenFog Reference Architecture for Fog Computing. Available online: https:\/\/www.iiconsortium.org\/pdf\/OpenFog_Reference_Architecture_2_09_17.pdf."},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Zhao, H., Zhang, W., Sun, H., and Xue, B. (2019). Embedded Deep Learning for Ship Detection and Recognition. Future Internet, 11.","DOI":"10.3390\/fi11020053"},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"3324","DOI":"10.1002\/sec.1537","article-title":"Modelling of Internet of Things units for estimating security-energy-performance relationships for quality of service and environment awareness","volume":"9","author":"Venckauskas","year":"2016","journal-title":"Secur. Commun. Netw."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"210","DOI":"10.1016\/j.comnet.2019.04.017","article-title":"Multi-sink distributed power control algorithm for Cyber-physical-systems in coal mine tunnels","volume":"161","author":"W","year":"2019","journal-title":"Comput. Netw."},{"key":"ref_26","unstructured":"Goodfellow, I., Bengio, Y., and Courville, A. (2016). Deep Learning, MIT Press."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"2278","DOI":"10.1109\/5.726791","article-title":"Gradient-based learning applied to document recognition","volume":"86","author":"Lecun","year":"1998","journal-title":"Proc IEEE"},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"21:1","DOI":"10.1145\/2735952","article-title":"A Real-Time Hand Posture Recognition System Using Deep Neural Networks","volume":"6","author":"Tang","year":"2015","journal-title":"ACM Trans. Intell. Syst. Technol."},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Wolf, M. (2017). Chapter 5\u2014Program Design and Analysis. Computers as Components, Morgan Kaufmann. [4th ed.].","DOI":"10.1016\/B978-0-12-805387-4.00005-4"},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"424","DOI":"10.1016\/j.future.2018.09.056","article-title":"Towards a practical framework for code offloading in the Internet of Things","volume":"92","author":"Benedetto","year":"2019","journal-title":"Future Gener. Comput. Syst."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"96","DOI":"10.1109\/MNET.2018.1700202","article-title":"Learning IoT in Edge: Deep Learning for the Internet of Things with Edge Computing","volume":"32","author":"Li","year":"2018","journal-title":"IEEE Netw."},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"2348","DOI":"10.1109\/TCAD.2018.2858384","article-title":"DeepThings: Distributed Adaptive Deep Learning Inference on Resource-Constrained IoT Edge Clusters","volume":"37","author":"Zhao","year":"2018","journal-title":"IEEE Trans. Comput. Aided Des. Integr. Circuits Syst."},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"291","DOI":"10.1002\/j.1538-7305.1970.tb01770.x","article-title":"An efficient heuristic procedure for partitioning graphs","volume":"49","author":"Kernighan","year":"1970","journal-title":"Bell Syst. Tech. J."},{"key":"ref_34","doi-asserted-by":"crossref","unstructured":"Al-Arnaout, Z., Hart, J., Fu, Q., and Frean, M. (2012, January 22\u201325). MP-DNA: A novel distributed replica placement heuristic for WMNs. Proceedings of the 37th Annual IEEE Conference on Local Computer Networks, Clearwater, FL, USA.","DOI":"10.1109\/LCN.2012.6423679"},{"key":"ref_35","doi-asserted-by":"crossref","unstructured":"Wen, X., Chen, K., Chen, Y., Liu, Y., Xia, Y., and Hu, C. (2012, January 18\u201321). VirtualKnotter: Online Virtual Machine Shuffling for Congestion Resolving in Virtualized Datacenter. Proceedings of the IEEE 32nd International Conference on Distributed Computing Systems Workshop, Macau, China.","DOI":"10.1109\/ICDCS.2012.25"},{"key":"ref_36","doi-asserted-by":"crossref","unstructured":"Cao, B., Gao, X., Chen, G., and Jin, Y. (2014, January 16\u201319). NICE: Network-aware VM Consolidation scheme for Energy Conservation in Data Centers. Proceedings of the 20th IEEE International Conference on Parallel and Distributed Systems (ICPADS), Hsinchu, Taiwan.","DOI":"10.1109\/PADSW.2014.7097805"},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"451","DOI":"10.1016\/j.future.2012.07.003","article-title":"Graph partitioning algorithms for optimizing software deployment in mobile cloud computing","volume":"29","author":"Verbelen","year":"2013","journal-title":"Future Gener. Comput. Syst."},{"key":"ref_38","unstructured":"Guerrieri, A., and Montresor, A. (2014). Distributed Edge Partitioning for Graph Processing. arXiv."},{"key":"ref_39","doi-asserted-by":"crossref","unstructured":"Magoutis, K., and Pietzuch, P. (2014). Distributed Vertex-Cut Partitioning. Lecture Notes in Computer Science, Proceedings of the Distributed Applications and Interoperable Systems, Berlin, Germany, 3\u20135 June 2014, Springer.","DOI":"10.1007\/978-3-662-43352-2"},{"key":"ref_40","doi-asserted-by":"crossref","first-page":"131","DOI":"10.1016\/j.ecoinf.2019.05.005","article-title":"Columnar cactus recognition in aerial images using a deep learning approach","volume":"52","year":"2019","journal-title":"Ecol. Inform."},{"key":"ref_41","doi-asserted-by":"crossref","first-page":"7500","DOI":"10.1080\/01431161.2019.1569282","article-title":"Young and mature oil palm tree detection and counting using convolutional neural network deep learning method","volume":"40","author":"Nadarajoo","year":"2019","journal-title":"Int. J. Remote Sens."},{"key":"ref_42","doi-asserted-by":"crossref","first-page":"524","DOI":"10.1016\/j.future.2018.11.036","article-title":"Background classification method based on deep learning for intelligent automotive radar target detection","volume":"94","author":"Ningbo","year":"2019","journal-title":"Future Gener. Comput. Syst."},{"key":"ref_43","unstructured":"STMicroelectronics (2019, July 24). STM32F469xx. Available online: https:\/\/www.st.com\/resource\/en\/datasheet\/stm32f469ae.pdf."},{"key":"ref_44","unstructured":"Atmel (2019, July 24). Atmel SAM G55G. Available online: http:\/\/ww1.microchip.com\/downloads\/en\/devicedoc\/Atmel-11289-32-bit-Cortex-M4-Microcontroller-SAM-G55_Datasheet.pdf."},{"key":"ref_45","unstructured":"STMicroelectronics (2019, July 24). STM32L433xx. Available online: https:\/\/www.st.com\/resource\/en\/datasheet\/stm32l433cc.pdf."},{"key":"ref_46","unstructured":"STMicroelectronics (2019, July 24). STM32L151x6\/8\/B. Available online: https:\/\/www.st.com\/resource\/en\/datasheet\/stm32l151vb.pdf."},{"key":"ref_47","unstructured":"Karypis, G. (2019, March 30). METIS A Software Package for Partitioning Unstructured Graphs, Partitioning Meshes, and Computing Fill-Reducing Orderings of Sparse Matrices Version 5.1.0. Available online: http:\/\/glaros.dtc.umn.edu\/gkhome\/fetch\/sw\/metis\/manual.pdf."},{"key":"ref_48","doi-asserted-by":"crossref","unstructured":"Honovich, J. (2019, July 14). Frame Rate Guide for Video Surveillance. Available online: https:\/\/ipvm.com\/reports\/frame-rate-surveillance-guide.","DOI":"10.4324\/9780429468872-2"},{"key":"ref_49","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1007\/BF01759032","article-title":"Retiming synchronous circuitry","volume":"6","author":"Leiserson","year":"1991","journal-title":"Algorithmica"},{"key":"ref_50","doi-asserted-by":"crossref","first-page":"208","DOI":"10.1007\/s001580100138","article-title":"Pareto analysis in multiobjective optimization using the collinearity theorem and scaling method","volume":"22","author":"Kasprzak","year":"2001","journal-title":"Struct. Multidiscip. Optim."}],"container-title":["Future Internet"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1999-5903\/11\/10\/209\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T13:25:55Z","timestamp":1760189155000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1999-5903\/11\/10\/209"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2019,9,29]]},"references-count":50,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2019,10]]}},"alternative-id":["fi11100209"],"URL":"https:\/\/doi.org\/10.3390\/fi11100209","relation":{},"ISSN":["1999-5903"],"issn-type":[{"value":"1999-5903","type":"electronic"}],"subject":[],"published":{"date-parts":[[2019,9,29]]}}}