Files in this item
ScissionLite: accelerating distributed deep learning with lightweight data compression for IIoT
Item metadata
dc.contributor.author | Ahn, Hyunho | |
dc.contributor.author | Lee, Munkyu | |
dc.contributor.author | Seong, Sihoon | |
dc.contributor.author | Na, Gap-Joo | |
dc.contributor.author | Chun, In-Geol | |
dc.contributor.author | Varghese, Blesson | |
dc.contributor.author | Hong, Cheol-Ho | |
dc.date.accessioned | 2024-07-10T10:30:13Z | |
dc.date.available | 2024-07-10T10:30:13Z | |
dc.date.issued | 2024-06-24 | |
dc.identifier | 302403613 | |
dc.identifier | 3ff2ab96-02af-4274-add5-9e6c4d8d7acd | |
dc.identifier.citation | Ahn , H , Lee , M , Seong , S , Na , G-J , Chun , I-G , Varghese , B & Hong , C-H 2024 , ' ScissionLite: accelerating distributed deep learning with lightweight data compression for IIoT ' , IEEE Transactions on Industrial Informatics , vol. Early Access . https://doi.org/10.1109/TII.2024.3413340 | en |
dc.identifier.issn | 1551-3203 | |
dc.identifier.uri | https://hdl.handle.net/10023/30140 | |
dc.description | Funding: This work was supported in part by the Electronics and Telecommunications Research Institute through the Korean government under Grant 23zs1300 (Research on High Performance Computing Technology to overcome limitations of AI processing) and in part by the Korea Institute for Advancement of Technology (KIAT) through the Korea Government (MOTIE) under Grant P0017011 (HRD Program for Industrial Innovation). Paper no. TII-23-4829. | en |
dc.description.abstract | Industrial Internet of Things (IIoT) applications can greatly benefit from leveraging edge computing. For instance, applications relying on deep neural network (DNN) models can be sliced and distributed across IIoT devices and the network edge to reduce inference latency. However, low network performance between IIoT devices and the edge often becomes a bottleneck. In this study, we propose ScissionLite, a holistic framework designed to accelerate distributed DNN inference using lightweight data compression. Our compression method features a novel lightweight down/upsampling network tailored for performance-limited IIoT devices, which is inserted at the slicing point of a DNN model to reduce outbound network traffic without causing a significant drop in accuracy. In addition, we have developed a benchmarking tool to accurately identify the optimal slicing point of the DNN for the best inference latency. ScissionLite improves inference latency by up to 15.7× with minimal accuracy degradation. | |
dc.format.extent | 11 | |
dc.format.extent | 1364664 | |
dc.language.iso | eng | |
dc.relation.ispartof | IEEE Transactions on Industrial Informatics | en |
dc.subject | Edge computing | en |
dc.subject | IIoT | en |
dc.subject | Deep neural networks | en |
dc.subject | Model slicing | en |
dc.subject | Inference | en |
dc.subject | QA75 Electronic computers. Computer science | en |
dc.subject | 3rd-NDAS | en |
dc.subject.lcc | QA75 | en |
dc.title | ScissionLite: accelerating distributed deep learning with lightweight data compression for IIoT | en |
dc.type | Journal article | en |
dc.contributor.institution | University of St Andrews. School of Computer Science | en |
dc.identifier.doi | 10.1109/TII.2024.3413340 | |
dc.description.status | Peer reviewed | en |
dc.date.embargoedUntil | 2024-06-24 |
This item appears in the following Collection(s)
Items in the St Andrews Research Repository are protected by copyright, with all rights reserved, unless otherwise indicated.