Deep Learning With Edge Computing: A Review

Deep learning is currently widely used in a variety of applications, including computer vision and natural language processing. End devices, such as smartphones and Internet-of-Things sensors, are generating data that need to be analyzed in real time using deep learning or used to train deep learnin...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Veröffentlicht in:	Proceedings of the IEEE Jg. 107; H. 8; S. 1655 - 1674
Hauptverfasser:	Chen, Jiasi, Ran, Xukan
Format:	Journal Article
Sprache:	Englisch
Veröffentlicht:	New York IEEE 01.08.2019 The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Schlagworte:	Artificial intelligence Cloud computing Computational modeling Computer vision Deep learning Devices Edge computing Finite element method Inference Machine learning mobile computing Natural language processing Neural networks Privacy Servers Smartphones State-of-the-art reviews Training
ISSN:	0018-9219, 1558-2256
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Abstract	Deep learning is currently widely used in a variety of applications, including computer vision and natural language processing. End devices, such as smartphones and Internet-of-Things sensors, are generating data that need to be analyzed in real time using deep learning or used to train deep learning models. However, deep learning inference and training require substantial computation resources to run quickly. Edge computing, where a fine mesh of compute nodes are placed close to end devices, is a viable way to meet the high computation and low-latency requirements of deep learning on edge devices and also provides additional benefits in terms of privacy, bandwidth efficiency, and scalability. This paper aims to provide a comprehensive review of the current state of the art at the intersection of deep learning and edge computing. Specifically, it will provide an overview of applications where deep learning is used at the network edge, discuss various approaches for quickly executing deep learning inference across a combination of end devices, edge servers, and the cloud, and describe the methods for training deep learning models across multiple edge devices. It will also discuss open challenges in terms of systems performance, network technologies and management, benchmarks, and privacy. The reader will take away the following concepts from this paper: understanding scenarios where deep learning at the network edge can be useful, understanding common techniques for speeding up deep learning inference and performing distributed training on edge devices, and understanding recent trends and opportunities.
AbstractList	Deep learning is currently widely used in a variety of applications, including computer vision and natural language processing. End devices, such as smartphones and Internet-of-Things sensors, are generating data that need to be analyzed in real time using deep learning or used to train deep learning models. However, deep learning inference and training require substantial computation resources to run quickly. Edge computing, where a fine mesh of compute nodes are placed close to end devices, is a viable way to meet the high computation and low-latency requirements of deep learning on edge devices and also provides additional benefits in terms of privacy, bandwidth efficiency, and scalability. This paper aims to provide a comprehensive review of the current state of the art at the intersection of deep learning and edge computing. Specifically, it will provide an overview of applications where deep learning is used at the network edge, discuss various approaches for quickly executing deep learning inference across a combination of end devices, edge servers, and the cloud, and describe the methods for training deep learning models across multiple edge devices. It will also discuss open challenges in terms of systems performance, network technologies and management, benchmarks, and privacy. The reader will take away the following concepts from this paper: understanding scenarios where deep learning at the network edge can be useful, understanding common techniques for speeding up deep learning inference and performing distributed training on edge devices, and understanding recent trends and opportunities.
Author	Chen, Jiasi Ran, Xukan
Author_xml	– sequence: 1 givenname: Jiasi orcidid: 0000-0001-9923-9027 surname: Chen fullname: Chen, Jiasi email: jiasi@cs.ucr.edu organization: Department of Computer Science and Engineering, University of California at Riverside, Riverside, CA, USA – sequence: 2 givenname: Xukan surname: Ran fullname: Ran, Xukan organization: Department of Computer Science and Engineering, University of California at Riverside, Riverside, CA, USA
BookMark	eNp9kE1Lw0AQhhepYFv9A3oJeJTU_cgku95KWr8oVIricdkkk7qlTeJmq_jvTW3x4MHTwMz7zDDPgPSqukJCzhkdMUbV9ePTYp6OOGVqxBVnKkmOSJ8ByJBziHukTymT4W5yQgZtu6KUCohFn1xNEJtghsZVtloGr9a_BdNiiUFab5qt73o3wThY4IfFz1NyXJp1i2eHOiQvt9Pn9D6cze8e0vEszLkCHxZlwgRkIJKCQqkizGOGWZQBFArj2AiTq0hIA5QrGWUouOGgIIuNoYXMSjEkl_u9javft9h6vaq3rupOas6T7hGhFO9Scp_KXd22DkudW2-8rSvvjF1rRvVOjf5Ro3dq9EFNh_I_aOPsxriv_6GLPWQR8ReQSSykBPENGytvTw
CODEN	IEEPAD
CitedBy_id	crossref_primary_10_1007_s10586_024_04413_7 crossref_primary_10_1109_JIOT_2024_3462722 crossref_primary_10_1177_16878132251347390 crossref_primary_10_1016_j_procs_2025_07_180 crossref_primary_10_1109_TSMC_2023_3297771 crossref_primary_10_3390_electronics11244157 crossref_primary_10_1016_j_comcom_2020_05_035 crossref_primary_10_1029_2024WR038931 crossref_primary_10_3390_agriculture12071033 crossref_primary_10_1109_ACCESS_2019_2950287 crossref_primary_10_3390_s24144589 crossref_primary_10_1109_TGCN_2022_3186879 crossref_primary_10_3390_app14093876 crossref_primary_10_1109_TBCAS_2022_3181808 crossref_primary_10_1145_3597936 crossref_primary_10_3390_math11041055 crossref_primary_10_1109_TVCG_2025_3574194 crossref_primary_10_1109_MNET_2024_3520166 crossref_primary_10_1109_TNNLS_2022_3154443 crossref_primary_10_1145_3543069 crossref_primary_10_3390_mi14071367 crossref_primary_10_1002_ett_4458 crossref_primary_10_1109_TCC_2022_3177649 crossref_primary_10_1109_JSYST_2020_2991814 crossref_primary_10_1038_s41467_022_29171_1 crossref_primary_10_1016_j_procir_2020_03_081 crossref_primary_10_1007_s00607_021_00953_7 crossref_primary_10_1109_ACCESS_2023_3287093 crossref_primary_10_1002_spe_3301 crossref_primary_10_1088_1742_6596_2456_1_012035 crossref_primary_10_3389_fpsyg_2022_843493 crossref_primary_10_3390_electronics11182939 crossref_primary_10_1016_j_measen_2024_101177 crossref_primary_10_1109_TCSI_2023_3329337 crossref_primary_10_1145_3594539 crossref_primary_10_1007_s11517_022_02696_9 crossref_primary_10_1016_j_actaastro_2023_08_001 crossref_primary_10_1016_j_asoc_2021_107162 crossref_primary_10_1109_TCOMM_2024_3486979 crossref_primary_10_1109_TMC_2022_3154644 crossref_primary_10_1002_adma_202303481 crossref_primary_10_1109_ACCESS_2022_3165744 crossref_primary_10_1109_TMM_2023_3348662 crossref_primary_10_3390_s23041911 crossref_primary_10_1109_ACCESS_2025_3587626 crossref_primary_10_1093_comjnl_bxab188 crossref_primary_10_1007_s00521_023_08712_9 crossref_primary_10_1016_j_comcom_2021_07_020 crossref_primary_10_1016_j_future_2022_03_030 crossref_primary_10_1016_j_compag_2024_109032 crossref_primary_10_1016_j_iot_2025_101686 crossref_primary_10_1109_JIOT_2021_3067382 crossref_primary_10_3390_agronomy14091878 crossref_primary_10_1109_OJCOMS_2024_3522103 crossref_primary_10_1109_TVT_2022_3219058 crossref_primary_10_1016_j_future_2020_06_007 crossref_primary_10_1109_JSEN_2023_3313155 crossref_primary_10_1016_j_sysarc_2024_103181 crossref_primary_10_1016_j_comnet_2024_110823 crossref_primary_10_1016_j_neucom_2025_130835 crossref_primary_10_1007_s11227_025_07799_2 crossref_primary_10_1109_JSAC_2022_3228558 crossref_primary_10_1007_s11265_023_01895_3 crossref_primary_10_1109_ACCESS_2022_3220694 crossref_primary_10_1109_TC_2020_3021199 crossref_primary_10_1109_TMC_2022_3177569 crossref_primary_10_1016_j_asoc_2022_109856 crossref_primary_10_1109_JSAC_2022_3227102 crossref_primary_10_1109_TWC_2023_3303232 crossref_primary_10_1016_j_envsoft_2025_106524 crossref_primary_10_3389_fnins_2024_1457623 crossref_primary_10_1186_s40537_020_00303_y crossref_primary_10_1002_cpe_7919 crossref_primary_10_3390_s20092533 crossref_primary_10_1109_TMC_2025_3539356 crossref_primary_10_1109_JIOT_2020_3004077 crossref_primary_10_3390_s20205796 crossref_primary_10_1109_JIOT_2023_3318647 crossref_primary_10_1016_j_mtsust_2025_101098 crossref_primary_10_23919_AISE_2005_000007 crossref_primary_10_1016_j_iot_2025_101677 crossref_primary_10_1088_1741_2552_abf521 crossref_primary_10_3390_electronics11152379 crossref_primary_10_1109_TMC_2024_3443260 crossref_primary_10_1109_TPDS_2023_3332333 crossref_primary_10_1109_IOTM_001_2200152 crossref_primary_10_1109_JEDS_2024_3508759 crossref_primary_10_1109_ACCESS_2021_3094741 crossref_primary_10_1109_JIOT_2024_3492066 crossref_primary_10_1109_JIOT_2025_3557408 crossref_primary_10_3390_jimaging10120326 crossref_primary_10_1109_TKDE_2024_3377710 crossref_primary_10_1093_nsr_nwae441 crossref_primary_10_1016_j_iot_2024_101143 crossref_primary_10_3390_electronics14122397 crossref_primary_10_1016_j_scs_2023_104815 crossref_primary_10_3390_s21144946 crossref_primary_10_1109_TCAD_2021_3075420 crossref_primary_10_1016_j_suscom_2023_100870 crossref_primary_10_1088_1361_6501_aca991 crossref_primary_10_1109_ACCESS_2023_3234761 crossref_primary_10_1016_j_neucom_2021_04_141 crossref_primary_10_1016_j_iot_2024_101158 crossref_primary_10_1109_TCSVT_2024_3466524 crossref_primary_10_1016_j_iot_2024_101272 crossref_primary_10_1109_JPROC_2022_3153408 crossref_primary_10_1109_TC_2023_3310678 crossref_primary_10_1587_transele_2024LHP0003 crossref_primary_10_1109_JSSC_2022_3140753 crossref_primary_10_1016_j_trc_2022_103982 crossref_primary_10_1109_TWC_2023_3241841 crossref_primary_10_1007_s10791_025_09509_1 crossref_primary_10_1016_j_snb_2022_132363 crossref_primary_10_1109_TCOMM_2024_3418901 crossref_primary_10_3390_s23094258 crossref_primary_10_1007_s10586_024_04893_7 crossref_primary_10_1016_j_future_2022_02_024 crossref_primary_10_1007_s11063_022_10951_1 crossref_primary_10_1109_COMST_2023_3319952 crossref_primary_10_1016_j_ifacol_2022_12_042 crossref_primary_10_1016_j_spc_2022_05_022 crossref_primary_10_1017_aer_2024_23 crossref_primary_10_1109_TETC_2021_3054761 crossref_primary_10_1007_s11356_022_19358_w crossref_primary_10_1016_j_procs_2024_09_443 crossref_primary_10_1016_j_image_2025_117306 crossref_primary_10_1109_TNNLS_2021_3104860 crossref_primary_10_1109_TPDS_2022_3177782 crossref_primary_10_1145_3698767 crossref_primary_10_3390_ijgi10020081 crossref_primary_10_1109_COMST_2023_3323091 crossref_primary_10_3389_fnmol_2022_999605 crossref_primary_10_1109_TIFS_2023_3312973 crossref_primary_10_1109_TFUZZ_2022_3163909 crossref_primary_10_1109_JIOT_2021_3131407 crossref_primary_10_1038_s41598_024_56522_3 crossref_primary_10_3390_en14123654 crossref_primary_10_1007_s11235_024_01185_8 crossref_primary_10_1109_ACCESS_2025_3577908 crossref_primary_10_1016_j_comnet_2022_109380 crossref_primary_10_1109_TMC_2024_3427709 crossref_primary_10_1109_JIOT_2025_3557431 crossref_primary_10_1109_ACCESS_2024_3351308 crossref_primary_10_3390_electronics11234007 crossref_primary_10_1109_COMST_2022_3200740 crossref_primary_10_1109_TGCN_2022_3208819 crossref_primary_10_1109_TMC_2025_3540017 crossref_primary_10_1016_j_sysarc_2021_102062 crossref_primary_10_1109_JIOT_2024_3446640 crossref_primary_10_34133_plantphenomics_0062 crossref_primary_10_1016_j_nexres_2025_100424 crossref_primary_10_3390_app122010619 crossref_primary_10_1016_j_comnet_2022_109032 crossref_primary_10_1109_JSEN_2023_3285618 crossref_primary_10_1109_TNNLS_2021_3082304 crossref_primary_10_1016_j_neunet_2025_108127 crossref_primary_10_1109_TII_2022_3163137 crossref_primary_10_1155_er_8824853 crossref_primary_10_1109_JSTSP_2023_3312914 crossref_primary_10_1016_j_rser_2021_111051 crossref_primary_10_11648_j_ajnc_20251402_13 crossref_primary_10_1039_D4MH01676A crossref_primary_10_3390_electronics14091802 crossref_primary_10_3390_fi16090329 crossref_primary_10_1142_S0219455424501608 crossref_primary_10_1016_j_engappai_2023_106407 crossref_primary_10_1109_ACCESS_2022_3210584 crossref_primary_10_3390_s22145405 crossref_primary_10_1016_j_sysarc_2021_102198 crossref_primary_10_1109_ACCESS_2021_3099311 crossref_primary_10_1109_TGCN_2021_3058657 crossref_primary_10_1049_ipr2_12449 crossref_primary_10_1109_JPROC_2022_3223186 crossref_primary_10_3390_fire6080315 crossref_primary_10_3390_math11010156 crossref_primary_10_1007_s10009_023_00695_1 crossref_primary_10_1016_j_entcom_2024_100781 crossref_primary_10_1039_D3MH01253K crossref_primary_10_1109_ACCESS_2023_3265357 crossref_primary_10_3390_s22093208 crossref_primary_10_1140_epjb_s10051_024_00703_6 crossref_primary_10_1109_JPROC_2021_3056006 crossref_primary_10_1109_JIOT_2025_3555153 crossref_primary_10_1109_LGRS_2021_3072028 crossref_primary_10_1109_TMC_2024_3357218 crossref_primary_10_1016_j_neucom_2021_01_122 crossref_primary_10_1145_3512798_3512821 crossref_primary_10_1088_2634_4386_ad8c78 crossref_primary_10_3390_s24103098 crossref_primary_10_3390_electronics12051154 crossref_primary_10_1109_JIOT_2025_3584811 crossref_primary_10_3390_app112411807 crossref_primary_10_1016_j_displa_2024_102883 crossref_primary_10_1016_j_future_2022_01_013 crossref_primary_10_1016_j_inffus_2020_10_014 crossref_primary_10_1109_JIOT_2025_3566284 crossref_primary_10_1016_j_comnet_2025_111468 crossref_primary_10_1109_ACCESS_2024_3373691 crossref_primary_10_3390_electronics14122401 crossref_primary_10_1109_ACCESS_2021_3118731 crossref_primary_10_1109_ACCESS_2023_3268763 crossref_primary_10_1016_j_phycom_2024_102460 crossref_primary_10_3390_app15073863 crossref_primary_10_3390_s23177310 crossref_primary_10_3390_app11104607 crossref_primary_10_1016_j_sysarc_2020_101830 crossref_primary_10_1016_j_prevetmed_2024_106300 crossref_primary_10_1145_3464428 crossref_primary_10_3390_data10070116 crossref_primary_10_3390_s23052766 crossref_primary_10_1109_TMC_2024_3475634 crossref_primary_10_1016_j_ecoinf_2023_102065 crossref_primary_10_1109_JIOT_2020_3038848 crossref_primary_10_1063_5_0211040 crossref_primary_10_1007_s10015_021_00683_1 crossref_primary_10_1080_0951192X_2023_2189312 crossref_primary_10_1109_ACCESS_2023_3280411 crossref_primary_10_1109_ACCESS_2023_3314381 crossref_primary_10_1109_JSTARS_2025_3536175 crossref_primary_10_3390_info12040146 crossref_primary_10_3390_s21134412 crossref_primary_10_1016_j_compag_2022_107200 crossref_primary_10_1016_j_phycom_2022_101620 crossref_primary_10_1109_TCCN_2021_3103511 crossref_primary_10_1109_JPROC_2022_3226481 crossref_primary_10_1016_j_iot_2024_101375 crossref_primary_10_1109_JIOT_2020_2981338 crossref_primary_10_1007_s10462_024_11033_5 crossref_primary_10_1016_j_engappai_2025_110371 crossref_primary_10_3390_app15147738 crossref_primary_10_3390_pr13061917 crossref_primary_10_1109_COMST_2022_3199544 crossref_primary_10_1109_TMC_2024_3411295 crossref_primary_10_32628_CSEIT251112250 crossref_primary_10_1109_TCOMM_2021_3066495 crossref_primary_10_1109_ACCESS_2024_3477293 crossref_primary_10_1109_TC_2021_3135752 crossref_primary_10_3390_electronics10232958 crossref_primary_10_1016_j_ndteint_2022_102703 crossref_primary_10_1145_3486674 crossref_primary_10_1109_TETCI_2020_3007905 crossref_primary_10_3390_electronics11172700 crossref_primary_10_1109_JETCAS_2020_3040300 crossref_primary_10_1109_JPROC_2024_3437365 crossref_primary_10_1109_JIOT_2023_3333293 crossref_primary_10_1109_JSEN_2024_3502539 crossref_primary_10_1007_s12145_025_01863_4 crossref_primary_10_3390_rs13214370 crossref_primary_10_26599_TST_2023_9010086 crossref_primary_10_1109_TMC_2024_3418613 crossref_primary_10_1007_s43995_024_00060_6 crossref_primary_10_1038_s41746_024_01005_y crossref_primary_10_1109_ACCESS_2020_3000867 crossref_primary_10_1109_ACCESS_2021_3066559 crossref_primary_10_1007_s10586_024_04732_9 crossref_primary_10_1145_3464419 crossref_primary_10_3390_s25175302 crossref_primary_10_1155_2022_7863972 crossref_primary_10_1109_ACCESS_2020_3011705 crossref_primary_10_1007_s10723_023_09658_x crossref_primary_10_1109_TNNLS_2022_3190451 crossref_primary_10_1007_s11036_024_02317_9 crossref_primary_10_3390_en16104060 crossref_primary_10_1109_TMC_2022_3157957 crossref_primary_10_1145_3476990 crossref_primary_10_1145_3591867 crossref_primary_10_1109_JSAC_2021_3118432 crossref_primary_10_1109_TMC_2025_3564390 crossref_primary_10_3390_drones7100620 crossref_primary_10_3390_s22051854 crossref_primary_10_3390_fi14020054 crossref_primary_10_3390_sym14122524 crossref_primary_10_1016_j_jisa_2024_103786 crossref_primary_10_3390_app10134567 crossref_primary_10_3389_fnins_2022_1018006 crossref_primary_10_3389_frai_2025_1640549 crossref_primary_10_1002_cpe_8253 crossref_primary_10_1109_MWC_011_2300344 crossref_primary_10_1109_JIOT_2020_3020911 crossref_primary_10_1109_ACCESS_2024_3487073 crossref_primary_10_1007_s10586_025_05386_x crossref_primary_10_1177_16878132241281470 crossref_primary_10_7717_peerj_cs_2306 crossref_primary_10_1007_s10845_025_02679_1 crossref_primary_10_3390_nano14070562 crossref_primary_10_1088_1361_6579_ad37ee crossref_primary_10_1098_rsos_230806 crossref_primary_10_1109_JIOT_2024_3486351 crossref_primary_10_1109_JSAC_2021_3118403 crossref_primary_10_1109_TC_2021_3107196 crossref_primary_10_3390_fi13010005 crossref_primary_10_1007_s00521_024_10718_w crossref_primary_10_1109_TSUSC_2025_3542563 crossref_primary_10_1109_TCSS_2024_3395467 crossref_primary_10_3390_rs13214387 crossref_primary_10_1109_TCSII_2022_3224022 crossref_primary_10_1016_j_ipm_2023_103544 crossref_primary_10_1109_ACCESS_2021_3074180 crossref_primary_10_1002_cpe_8002 crossref_primary_10_1007_s11432_024_4205_8 crossref_primary_10_1109_ACCESS_2024_3458420 crossref_primary_10_1109_ACCESS_2021_3124029 crossref_primary_10_1109_TMC_2021_3114193 crossref_primary_10_1109_MCE_2024_3506502 crossref_primary_10_1109_OJCOMS_2025_3555947 crossref_primary_10_1016_j_future_2021_07_031 crossref_primary_10_1109_TPDS_2024_3506588 crossref_primary_10_3390_agronomy15010242 crossref_primary_10_1109_ACCESS_2020_3029127 crossref_primary_10_1016_j_ultras_2025_107598 crossref_primary_10_1016_j_csi_2021_103521 crossref_primary_10_3390_en13164208 crossref_primary_10_1109_TSC_2025_3576692 crossref_primary_10_1109_LRA_2022_3219306 crossref_primary_10_1016_j_adhoc_2021_102727 crossref_primary_10_1145_3570604 crossref_primary_10_1038_s41598_025_90764_z crossref_primary_10_1016_j_jisa_2023_103483 crossref_primary_10_1109_TSP_2024_3351469 crossref_primary_10_1016_j_scs_2022_104059 crossref_primary_10_1109_ACCESS_2021_3059251 crossref_primary_10_1109_COMST_2020_2970550 crossref_primary_10_1016_j_aei_2024_102514 crossref_primary_10_1002_alr_23458 crossref_primary_10_1049_cdt2_6214436 crossref_primary_10_1007_s11277_024_11443_2 crossref_primary_10_1145_3461699 crossref_primary_10_2139_ssrn_5052124 crossref_primary_10_1016_j_future_2024_03_014 crossref_primary_10_1109_JSEN_2023_3347616 crossref_primary_10_1007_s00607_025_01418_x crossref_primary_10_1016_j_asoc_2020_106582 crossref_primary_10_1109_ACCESS_2024_3454812 crossref_primary_10_1109_JIOT_2024_3365957 crossref_primary_10_1109_TMC_2022_3172402 crossref_primary_10_3390_electronics10161912 crossref_primary_10_3390_s21041064 crossref_primary_10_1109_ACCESS_2024_3475996 crossref_primary_10_1109_ACCESS_2024_3349978 crossref_primary_10_1016_j_comcom_2023_06_010 crossref_primary_10_1109_JIOT_2023_3263290 crossref_primary_10_1109_JIOT_2021_3102945 crossref_primary_10_3390_s22197382 crossref_primary_10_1109_TWC_2023_3335302 crossref_primary_10_3788_AI_2025_10003 crossref_primary_10_1016_j_smhl_2021_100249 crossref_primary_10_1145_3527155 crossref_primary_10_1145_3679201 crossref_primary_10_1109_ACCESS_2020_2982411 crossref_primary_10_1109_COMST_2020_3035427 crossref_primary_10_1016_j_procs_2025_03_273 crossref_primary_10_1038_s41598_024_82272_3 crossref_primary_10_1007_s11042_023_14383_4 crossref_primary_10_1109_ACCESS_2021_3084689 crossref_primary_10_1109_TSC_2021_3109094 crossref_primary_10_1109_JIOT_2025_3527750 crossref_primary_10_1109_TGCN_2024_3373911 crossref_primary_10_3390_s25154586 crossref_primary_10_1109_LGRS_2021_3055795 crossref_primary_10_1016_j_engappai_2024_109475 crossref_primary_10_1007_s10723_023_09730_6 crossref_primary_10_12688_digitaltwin_17632_2 crossref_primary_10_32604_cmes_2024_048932 crossref_primary_10_1109_ACCESS_2020_3037717 crossref_primary_10_3390_s21155082 crossref_primary_10_3390_s24175633 crossref_primary_10_2478_eces_2023_0029 crossref_primary_10_1109_TCCN_2021_3056707 crossref_primary_10_3390_app15137533 crossref_primary_10_38124_ijisrt_25jul835 crossref_primary_10_1016_j_procs_2022_12_267 crossref_primary_10_1109_MWC_001_2000466 crossref_primary_10_1145_3439602_3439618 crossref_primary_10_3390_technologies12110212 crossref_primary_10_3390_wevj13060101 crossref_primary_10_1016_j_adhoc_2023_103156 crossref_primary_10_3390_math13050887 crossref_primary_10_1016_j_neunet_2022_06_038 crossref_primary_10_3390_app15042127 crossref_primary_10_1007_s12008_024_01943_7 crossref_primary_10_1109_TON_2024_3522073 crossref_primary_10_3390_app12157679 crossref_primary_10_1007_s10462_025_11346_z crossref_primary_10_1109_JIOT_2024_3432975 crossref_primary_10_1109_TCOMM_2023_3338738 crossref_primary_10_1109_TII_2023_3324939 crossref_primary_10_1109_TSP_2021_3090323 crossref_primary_10_1016_j_compchemeng_2021_107508 crossref_primary_10_1016_j_iot_2020_100314 crossref_primary_10_1109_MNET_011_1900450 crossref_primary_10_3389_fdata_2022_787421 crossref_primary_10_1016_j_comcom_2024_01_031 crossref_primary_10_1109_TWC_2024_3523517 crossref_primary_10_1109_JIOT_2023_3313514 crossref_primary_10_3390_s21134496 crossref_primary_10_1016_j_vlsi_2020_05_002 crossref_primary_10_1007_s11518_024_5604_1 crossref_primary_10_1016_j_vlsi_2023_102127 crossref_primary_10_1016_j_cej_2024_153382 crossref_primary_10_1109_MSP_2021_3125282 crossref_primary_10_1109_JIOT_2021_3112715 crossref_primary_10_3390_electronics11172678 crossref_primary_10_1016_j_compeleceng_2024_109202 crossref_primary_10_1016_j_eswa_2022_117823 crossref_primary_10_1080_15567036_2022_2119309 crossref_primary_10_38124_ijsrmt_v3i6_677 crossref_primary_10_1016_j_neucom_2023_02_011 crossref_primary_10_1016_j_resconrec_2025_108311 crossref_primary_10_3390_aerospace10010078 crossref_primary_10_1109_TNSE_2021_3055835 crossref_primary_10_3103_S8756699021020047 crossref_primary_10_3390_technologies12060081 crossref_primary_10_1109_ACCESS_2023_3267964 crossref_primary_10_1016_j_jpdc_2022_06_006 crossref_primary_10_1109_TRPMS_2025_3530774 crossref_primary_10_1016_j_buildenv_2024_112217 crossref_primary_10_1109_ACCESS_2022_3158319 crossref_primary_10_1109_LCOMM_2020_3032517 crossref_primary_10_3390_s23031522 crossref_primary_10_1016_j_cor_2023_106294 crossref_primary_10_1016_j_egyr_2025_08_049 crossref_primary_10_3390_jcm12144830 crossref_primary_10_1088_1361_6552_ac2670 crossref_primary_10_1007_s13132_024_02087_5 crossref_primary_10_3390_electronics14142796 crossref_primary_10_1007_s11554_021_01102_1 crossref_primary_10_1145_3469029 crossref_primary_10_46810_tdfd_1529139 crossref_primary_10_1109_JIOT_2021_3088875 crossref_primary_10_1109_ACCESS_2025_3555201 crossref_primary_10_1016_j_engappai_2020_103840 crossref_primary_10_1016_j_ultras_2022_106776 crossref_primary_10_1016_j_optlastec_2024_111766 crossref_primary_10_1109_JPROC_2023_3247480 crossref_primary_10_1145_3582080 crossref_primary_10_1109_TWC_2024_3497167 crossref_primary_10_1016_j_ins_2024_121265 crossref_primary_10_3390_a13050125 crossref_primary_10_1109_OJCAS_2025_3546067 crossref_primary_10_1007_s11227_023_05663_9 crossref_primary_10_1109_TPAMI_2025_3570545 crossref_primary_10_1364_JOCN_498951 crossref_primary_10_1109_ACCESS_2023_3263392 crossref_primary_10_1016_j_robot_2021_103780 crossref_primary_10_1145_3527169 crossref_primary_10_1109_JIOT_2020_2983660 crossref_primary_10_1016_j_future_2024_107600 crossref_primary_10_1016_j_neucom_2025_130485 crossref_primary_10_1109_ACCESS_2025_3530297 crossref_primary_10_1109_JIOT_2023_3268771 crossref_primary_10_1016_j_cosrev_2024_100668 crossref_primary_10_3389_fnins_2022_937782 crossref_primary_10_1109_JSAC_2023_3242730 crossref_primary_10_1162_neco_a_01518 crossref_primary_10_3390_math12142263 crossref_primary_10_1109_TPDS_2021_3135441 crossref_primary_10_1038_s41598_024_72125_4 crossref_primary_10_1109_TVT_2024_3444815 crossref_primary_10_1109_ACCESS_2021_3109733 crossref_primary_10_3390_electronics13244953 crossref_primary_10_1007_s11432_023_3957_4 crossref_primary_10_1109_JIOT_2023_3280746 crossref_primary_10_3390_fi16100365 crossref_primary_10_3390_app13137872 crossref_primary_10_1145_3612918 crossref_primary_10_1016_j_comnet_2021_108327 crossref_primary_10_1016_j_future_2025_107914 crossref_primary_10_3390_app13084982 crossref_primary_10_1088_1742_6596_1646_1_012016 crossref_primary_10_1126_sciadv_adu4323 crossref_primary_10_3390_iot4040021 crossref_primary_10_1109_ACCESS_2024_3440631 crossref_primary_10_1109_TCOMM_2024_3476073 crossref_primary_10_1109_TIM_2022_3205644 crossref_primary_10_1109_TMC_2023_3267805 crossref_primary_10_1109_ACCESS_2025_3544107 crossref_primary_10_1016_j_adhoc_2022_103044 crossref_primary_10_1088_2631_7990_ad94b8 crossref_primary_10_1016_j_adhoc_2020_102152 crossref_primary_10_1109_TWC_2022_3183632 crossref_primary_10_1016_j_fmre_2025_09_014 crossref_primary_10_1109_JIOT_2020_2994200 crossref_primary_10_3390_info14060318 crossref_primary_10_1109_JIOT_2022_3162581 crossref_primary_10_1109_MIS_2022_3214614 crossref_primary_10_1145_3542819 crossref_primary_10_3390_s21238071 crossref_primary_10_1016_j_sftr_2025_100761 crossref_primary_10_3390_s22176710 crossref_primary_10_3390_jlpea12020028 crossref_primary_10_1109_ACCESS_2021_3069210 crossref_primary_10_3389_fendo_2024_1338743 crossref_primary_10_1016_j_jpdc_2023_05_001 crossref_primary_10_1016_j_comnet_2024_110641 crossref_primary_10_1109_TIM_2024_3351239 crossref_primary_10_1145_3546192 crossref_primary_10_1109_LSENS_2023_3271988 crossref_primary_10_3390_s23063330 crossref_primary_10_1109_TCC_2022_3160129 crossref_primary_10_1002_aelm_202200378 crossref_primary_10_1109_TSC_2022_3155447 crossref_primary_10_1007_s11276_019_02181_6 crossref_primary_10_1016_j_vlsi_2025_102508 crossref_primary_10_1145_3506718 crossref_primary_10_1145_3593044 crossref_primary_10_1109_JBHI_2020_3007661 crossref_primary_10_1007_s42514_022_00091_2 crossref_primary_10_1186_s13638_022_02152_0 crossref_primary_10_1109_COMST_2020_2988367 crossref_primary_10_1016_j_neucom_2021_09_045 crossref_primary_10_1007_s11227_022_04463_x crossref_primary_10_1016_j_compag_2024_108792 crossref_primary_10_1155_2020_6661022 crossref_primary_10_3390_electronics13142846 crossref_primary_10_1016_j_tifs_2025_105017 crossref_primary_10_1109_ACCESS_2022_3178999 crossref_primary_10_1109_JSTARS_2024_3389641 crossref_primary_10_1109_TNSE_2021_3054244 crossref_primary_10_1145_3665898 crossref_primary_10_1109_JIOT_2021_3107431 crossref_primary_10_1109_ACCESS_2023_3322370 crossref_primary_10_1109_TCC_2023_3258982 crossref_primary_10_1145_3448416 crossref_primary_10_1109_JPROC_2020_3033753 crossref_primary_10_1145_3503954_3503956 crossref_primary_10_1145_3494998 crossref_primary_10_1186_s13677_023_00493_9 crossref_primary_10_1109_COMST_2021_3135829 crossref_primary_10_1109_ACCESS_2021_3131396 crossref_primary_10_3390_pr13041144 crossref_primary_10_1016_j_trd_2022_103439 crossref_primary_10_1109_TCOMM_2022_3229033 crossref_primary_10_1016_j_atech_2025_101330 crossref_primary_10_3390_s23031388 crossref_primary_10_1038_s41598_023_31694_6 crossref_primary_10_1587_nolta_12_412 crossref_primary_10_1109_JETCAS_2020_3033155 crossref_primary_10_3390_jsan13060080 crossref_primary_10_58564_IJSER_2_4_2023_125 crossref_primary_10_1109_ACCESS_2022_3190538 crossref_primary_10_1109_TCC_2022_3175725 crossref_primary_10_1109_ACCESS_2020_3037108 crossref_primary_10_1109_ACCESS_2021_3081353 crossref_primary_10_1016_j_oceaneng_2024_117539 crossref_primary_10_1007_s11227_022_04819_3 crossref_primary_10_1016_j_future_2023_11_038 crossref_primary_10_1016_j_tej_2020_106883 crossref_primary_10_1080_00207543_2022_2058432 crossref_primary_10_1016_j_engappai_2024_109648 crossref_primary_10_1109_COMST_2023_3244674 crossref_primary_10_1109_JIOT_2025_3550162 crossref_primary_10_3788_PI_2025_C04 crossref_primary_10_3390_computers13110297 crossref_primary_10_1109_TNSE_2021_3054583 crossref_primary_10_1109_TVLSI_2025_3543445 crossref_primary_10_1007_s11063_022_11058_3 crossref_primary_10_1109_TKDE_2023_3269017 crossref_primary_10_1007_s11263_020_01379_y crossref_primary_10_1109_TCDS_2023_3275587 crossref_primary_10_1063_5_0237090 crossref_primary_10_1109_TSP_2020_3046971 crossref_primary_10_3390_ani12040437 crossref_primary_10_32604_cmc_2023_039487 crossref_primary_10_1016_j_jpdc_2022_04_004 crossref_primary_10_1109_JETCAS_2020_3015509 crossref_primary_10_1109_JIOT_2022_3182394 crossref_primary_10_1109_JLT_2025_3534275 crossref_primary_10_32604_cmes_2023_023684 crossref_primary_10_3233_XST_230429 crossref_primary_10_3390_electronics9010153 crossref_primary_10_1145_3530884 crossref_primary_10_1109_JIOT_2024_3439852 crossref_primary_10_1016_j_compag_2025_109908 crossref_primary_10_1016_j_teler_2024_100116 crossref_primary_10_1145_3764930 crossref_primary_10_1155_2022_2890473 crossref_primary_10_1109_TVLSI_2024_3375111 crossref_primary_10_1002_ett_4706 crossref_primary_10_1109_ACCESS_2022_3203394 crossref_primary_10_1016_j_psep_2025_107402 crossref_primary_10_1016_j_icte_2022_07_002 crossref_primary_10_3390_s21217298 crossref_primary_10_1016_j_ress_2023_109850 crossref_primary_10_1145_3555802 crossref_primary_10_3390_s21041031 crossref_primary_10_1007_s10922_023_09770_z crossref_primary_10_1109_COMST_2022_3218527 crossref_primary_10_1007_s11276_021_02554_w crossref_primary_10_1109_JSEN_2024_3469539 crossref_primary_10_1016_j_bspc_2023_105263 crossref_primary_10_1109_TNSE_2021_3067454 crossref_primary_10_3390_s23125672 crossref_primary_10_1186_s13635_023_00142_3 crossref_primary_10_1016_j_cose_2025_104455 crossref_primary_10_1109_TIFS_2023_3262149 crossref_primary_10_1016_j_cosrev_2020_100303 crossref_primary_10_1109_TSP_2023_3244092 crossref_primary_10_1109_ACCESS_2022_3202172 crossref_primary_10_1111_coin_12592 crossref_primary_10_1016_j_aej_2024_08_031 crossref_primary_10_4103_jpi_jpi_112_20 crossref_primary_10_1109_JPROC_2023_3306773 crossref_primary_10_1145_3479432 crossref_primary_10_1111_exsy_12836 crossref_primary_10_1007_s40747_021_00498_4 crossref_primary_10_1016_j_eswa_2024_124599 crossref_primary_10_1109_TNET_2021_3093452 crossref_primary_10_3390_electronics13152988 crossref_primary_10_1016_j_patcog_2021_108146 crossref_primary_10_1109_JIOT_2020_2983278 crossref_primary_10_1088_1741_2552_abf473 crossref_primary_10_1145_3524069 crossref_primary_10_1109_TGCN_2022_3146487 crossref_primary_10_1109_TMM_2024_3521704 crossref_primary_10_1016_j_sysarc_2022_102634 crossref_primary_10_1109_COMST_2024_3393230 crossref_primary_10_1109_JIOT_2023_3239944 crossref_primary_10_1016_j_comcom_2020_08_005 crossref_primary_10_1016_j_neunet_2023_10_006 crossref_primary_10_1155_dsn_5651009 crossref_primary_10_1007_s10586_021_03440_y crossref_primary_10_1109_ACCESS_2021_3050670 crossref_primary_10_1007_s13721_025_00600_7
Cites_doi	10.1145/3211332.3211336 10.1145/2733373.2807412 10.1109/COMST.2017.2745201 10.1109/CVPR.2017.351 10.1109/SEC.2018.00016 10.1109/NCA.2017.8171350 10.1145/2872887.2750389 10.1145/2647868.2654889 10.1145/3097895.3097903 10.1109/TETC.2014.2369958 10.1109/MCOM.2017.1700168 10.1109/INFOCOM.2018.8486403 10.1109/CVPR.2014.220 10.1145/2789168.2790123 10.1109/JIOT.2016.2579198 10.1109/TSC.2017.2662008 10.1145/3195970.3196023 10.1145/2999572.2999604 10.1145/3038912.3052577 10.1109/MICRO.2018.00023 10.1145/3230543.3230574 10.1145/3300061.3300116 10.1007/978-3-319-46448-0_2 10.1145/3020078.3021745 10.1109/MCI.2018.2840738 10.1145/3132211.3134445 10.1145/3081333.3081360 10.1145/3132211.3134460 10.1145/3132211.3134458 10.1145/3089801.3089805 10.1145/2594368.2594383 10.1007/s11263-015-0816-y 10.1145/3093336.3037698 10.1109/IPDPSW.2018.00098 10.1109/CVPR.2017.690 10.1109/ICDCS.2017.226 10.1145/2906388.2906396 10.1109/ICCV.2015.279 10.1109/TCAD.2018.2858384 10.1145/3196494.3196522 10.1109/IPSN.2016.7460664 10.1109/MNET.2018.1800109 10.1145/2462456.2464448 10.1145/3132211.3134459 10.1109/ICPR.2016.7900006 10.1145/3081333.3081347 10.1145/2994551.2994564 10.1109/MPRV.2009.82 10.1016/S0893-6080(98)00116-6 10.1145/2810103.2813687 10.1145/3229625.3229629 10.1109/COMST.2018.2844341 10.1109/CVPR.2018.00559 10.1145/3097895.3097896 10.1109/ICCV.2013.257 10.1162/neco.1997.9.8.1735 10.1109/MNET.2018.1700202 10.1109/MCOM.2018.1701277 10.1145/3240765.3243473 10.1109/SEC.2018.00014 10.1145/2976749.2978318 10.18653/v1/N16-1030 10.1109/JPROC.2017.2761740 10.1145/3133956.3134056 10.1145/2809695.2809711 10.1145/1999995.2000000 10.1007/11681878_14 10.1109/MC.2017.9 10.14722/ndss.2018.23204 10.1109/INFCOM.2012.6195689 10.1145/3219819.3220106 10.1145/2996864 10.1109/SP.2017.41 10.1109/SP.2017.12 10.23919/DATE.2017.7927211 10.1109/SEC.2016.38 10.1109/5.726791 10.1109/INFOCOM.2018.8485905 10.1145/2999572.2999602 10.1145/3123266.3123389 10.1145/1814433.1814441
ContentType	Journal Article
Copyright	Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2019
Copyright_xml	– notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2019
DBID	97E RIA RIE AAYXX CITATION 7SP 8FD L7M
DOI	10.1109/JPROC.2019.2921977
DatabaseName	IEEE All-Society Periodicals Package (ASPP) 2005–Present IEEE All-Society Periodicals Package (ASPP) 1998–Present IEEE Electronic Library (IEL) CrossRef Electronics & Communications Abstracts Technology Research Database Advanced Technologies Database with Aerospace
DatabaseTitle	CrossRef Technology Research Database Advanced Technologies Database with Aerospace Electronics & Communications Abstracts
DatabaseTitleList	Technology Research Database
Database_xml	– sequence: 1 dbid: RIE name: IEEE Electronic Library (IEL) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher
DeliveryMethod	fulltext_linktorsrc
Discipline	Engineering
EISSN	1558-2256
EndPage	1674
ExternalDocumentID	10_1109_JPROC_2019_2921977 8763885
Genre	orig-research
GrantInformation_xml	– fundername: National Science Foundation grantid: CNS-1817216 funderid: 10.13039/100000001
GroupedDBID	-DZ -~X .DC 0R~ 123 1OL 29P 3EH 4.4 6IK 85S 97E 9M8 AAJGR AAWTH ABAZT ABFSI ABJNI ABQJQ ABVLG ACBEA ACGFS AENEX AETEA AETIX AFOGA AGNAY AGQYO AGSQL AHBIQ AIBXA ALLEH ALMA_UNASSIGNED_HOLDINGS AZLTO BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 DU5 EBS EJD FA8 HZ~ H~9 IAAWW IBMZZ ICLAB IDIHD IFIPE IFJZH IPLJI JAVBF LAI M43 MVM O9- OCL RIA RIE RIU RNS TAE TN5 TWZ UDY UHB UKR UQL VOH WHG XJT XOL YNT ZCA ZXP ZY4 ~02 AAYXX CITATION 7SP 8FD L7M
ID	FETCH-LOGICAL-c295t-df7135b537d05f94ec61eb4b55d9e66a3ac9438a502984be32a2595b6aa0d8bf3
IEDL.DBID	RIE
ISICitedReferencesCount	916
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000497973300011&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	0018-9219
IngestDate	Sun Nov 30 03:55:52 EST 2025 Tue Nov 18 21:38:55 EST 2025 Sat Nov 29 06:01:42 EST 2025 Wed Aug 27 02:44:06 EDT 2025
IsPeerReviewed	true
IsScholarly	true
Issue	8
Language	English
License	https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html https://doi.org/10.15223/policy-029 https://doi.org/10.15223/policy-037
LinkModel	DirectLink
MergedId	FETCHMERGED-LOGICAL-c295t-df7135b537d05f94ec61eb4b55d9e66a3ac9438a502984be32a2595b6aa0d8bf3
Notes	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14
ORCID	0000-0001-9923-9027
PQID	2270183992
PQPubID	85453
PageCount	20
ParticipantIDs	ieee_primary_8763885 crossref_citationtrail_10_1109_JPROC_2019_2921977 proquest_journals_2270183992 crossref_primary_10_1109_JPROC_2019_2921977
PublicationCentury	2000
PublicationDate	2019-08-01
PublicationDateYYYYMMDD	2019-08-01
PublicationDate_xml	– month: 08 year: 2019 text: 2019-08-01 day: 01
PublicationDecade	2010
PublicationPlace	New York
PublicationPlace_xml	– name: New York
PublicationTitle	Proceedings of the IEEE
PublicationTitleAbbrev	JPROC
PublicationYear	2019
Publisher	IEEE The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml	– name: IEEE – name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References	ref57 ref56 ref58 ref53 ref52 (ref35) 0 zhang (ref124) 2015 goyal (ref139) 2017; abs 1706 2677 ref50 ruder (ref15) 2016; abs 1609 4747 (ref46) 2017 lin (ref129) 2017; abs 1712 1887 blot (ref131) 2016 (ref26) 0 iandola (ref70) 2016 gilad-bachrach (ref116) 2016 ref41 ref44 dean (ref121) 2012 jindal (ref65) 2018 ren (ref36) 2015 he (ref103) 2015; abs 1512 3385 ref8 (ref30) 0 ref9 ref4 ref101 szegedy (ref102) 2014; abs 1409 4842 zhong (ref55) 2017; abs 1712 8132 (ref43) 0 ref37 ref148 ref149 ref33 ref146 ref147 kusupati (ref47) 2018 (ref5) 0 ref39 ref38 ha (ref143) 2017 ovtcharov (ref82) 0 zhang (ref135) 2018 krizhevsky (ref99) 2012 (ref85) 0 (ref24) 0 dwork (ref150) 2006 ref23 saputra (ref54) 2018; abs 1812 5374 ref20 hsieh (ref127) 2017 (ref108) 0 ref27 (ref32) 0 anil (ref128) 2018; abs 1804 3235 zhang (ref93) 2017 (ref29) 0 papineni (ref16) 2002 jiang (ref95) 2018 ref13 liu (ref34) 2017 ref12 ref126 ref96 ref11 ref98 drolia (ref10) 2017 ref17 ref19 blog (ref6) 0 ref18 (ref31) 0 chinchali (ref51) 2018 simonyan (ref100) 2014; abs 1409 1556 jin (ref132) 2016 (ref83) 0 ref133 ref134 ref92 goodfellow (ref7) 2016; 1 ref94 mcmahan (ref125) 2016 (ref45) 0 ref130 loc huynh (ref78) 2017 ref91 ref90 ref89 ref137 ref138 ref88 ref136 ref87 (ref79) 0 teerapittayanon (ref110) 2017; abs 1709 1686 ref144 ref145 ref81 (ref86) 0 ref142 ref84 (ref48) 0 (ref40) 0 ref140 ref141 ref80 yao (ref77) 2018 kumar (ref59) 2017; 70 ref109 ref106 howard (ref21) 2017 ref107 han (ref71) 2015 lavalle (ref66) 2016 ref104 tan (ref22) 2018 ref74 ref105 ref76 ref2 ref1 huang (ref97) 2015; abs 1508 1991 chen (ref123) 2016; abs 1604 981 ref111 ref112 ref73 ref72 hinton (ref75) 2015 (ref28) 0 ref68 ref67 ref117 ref69 ref118 abadi (ref25) 2016; abs 1603 4467 ref64 ref115 ref63 ref113 ref114 jeans (ref3) 2019 ryan (ref49) 1998 ref60 ref122 ref62 ref120 ref61 wu (ref42) 2016 juvekar (ref119) 2018 shokri (ref14) 2015
References_xml	– ident: ref23 doi: 10.1145/3211332.3211336 – ident: ref101 doi: 10.1145/2733373.2807412 – year: 0 ident: ref6 publication-title: Edge Computing at Chick-Fil-A – year: 2016 ident: ref131 article-title: Gossip training for deep learning publication-title: arXiv 1611 09726 – ident: ref9 doi: 10.1109/COMST.2017.2745201 – ident: ref17 doi: 10.1109/CVPR.2017.351 – year: 0 ident: ref108 publication-title: Arcore Overview – ident: ref38 doi: 10.1109/SEC.2018.00016 – ident: ref130 doi: 10.1109/NCA.2017.8171350 – volume: 1 year: 2016 ident: ref7 publication-title: Deep Learning – ident: ref80 doi: 10.1145/2872887.2750389 – ident: ref27 doi: 10.1145/2647868.2654889 – ident: ref111 doi: 10.1145/3097895.3097903 – year: 2016 ident: ref66 publication-title: Virtual Reality – year: 0 ident: ref24 publication-title: TensorFlow™ – ident: ref52 doi: 10.1109/TETC.2014.2369958 – ident: ref58 doi: 10.1109/MCOM.2017.1700168 – ident: ref126 doi: 10.1109/INFOCOM.2018.8486403 – ident: ref105 doi: 10.1109/CVPR.2014.220 – ident: ref37 doi: 10.1145/2789168.2790123 – ident: ref8 doi: 10.1109/JIOT.2016.2579198 – year: 0 ident: ref86 publication-title: Qualcomm Neural Processing SDK for AI – start-page: 943 year: 1998 ident: ref49 article-title: Intrusion detection with neural networks publication-title: Proc Adv Neural Inf Process Syst – year: 0 ident: ref79 publication-title: Edge TPU – ident: ref92 doi: 10.1109/TSC.2017.2662008 – ident: ref118 doi: 10.1145/3195970.3196023 – ident: ref140 doi: 10.1109/NCA.2017.8171350 – year: 0 ident: ref35 publication-title: AWS Deeplens – year: 2018 ident: ref22 article-title: MnasNet: Platform-aware neural architecture search for mobile publication-title: arXiv 1807 11626 – ident: ref148 doi: 10.1145/2999572.2999604 – ident: ref56 doi: 10.1145/3038912.3052577 – year: 0 ident: ref43 publication-title: Google duplex An AI system for accomplishing real-world tasks over the phone – year: 0 ident: ref5 publication-title: AT&T Multi-Access Edge Computing – ident: ref133 doi: 10.1109/MICRO.2018.00023 – ident: ref94 doi: 10.1145/3230543.3230574 – ident: ref33 doi: 10.1145/3300061.3300116 – ident: ref69 doi: 10.1007/978-3-319-46448-0_2 – ident: ref73 doi: 10.1145/3020078.3021745 – ident: ref39 doi: 10.1109/MCI.2018.2840738 – ident: ref145 doi: 10.1145/3132211.3134445 – start-page: 311 year: 2002 ident: ref16 article-title: BLEU: A method for automatic evaluation of machine translation publication-title: Proc Annual Meeting of the Assoc Computational Linguistics – ident: ref12 doi: 10.1145/3081333.3081360 – year: 2016 ident: ref70 article-title: SqueezeNet: Alexnet-level accuracy with 50x fewer parameters and <0.5 MB model size publication-title: arXiv 1602 07360 – ident: ref144 doi: 10.1145/3132211.3134460 – volume: abs 1409 1556 start-page: 1 year: 2014 ident: ref100 article-title: Very deep convolutional networks for large-scale image recognition publication-title: CoRR – start-page: 1223 year: 2012 ident: ref121 article-title: Large scale distributed deep networks publication-title: Proc Adv Neural Inf Process Syst – ident: ref67 doi: 10.1145/3132211.3134458 – ident: ref87 doi: 10.1145/3089801.3089805 – start-page: 389 year: 2018 ident: ref77 article-title: On-demand deep model compression for mobile devices: A usage-driven model selection framework publication-title: Proc MobiSys – volume: abs 1706 2677 start-page: 1 year: 2017 ident: ref139 article-title: Accurate, large minibatch SGD: Training imagenet in 1 hour publication-title: CoRR – ident: ref68 doi: 10.1145/2594368.2594383 – ident: ref1 doi: 10.1007/s11263-015-0816-y – year: 2016 ident: ref125 article-title: Communication-efficient learning of deep networks from decentralized data publication-title: arXiv 1602 05629 – ident: ref13 doi: 10.1145/3093336.3037698 – ident: ref84 doi: 10.1109/IPDPSW.2018.00098 – ident: ref18 doi: 10.1109/CVPR.2017.690 – start-page: 1 year: 2018 ident: ref51 article-title: Cellular network traffic scheduling with deep reinforcement learning publication-title: Proc 32nd AAAI Conf Artif Intell – ident: ref109 doi: 10.1109/ICDCS.2017.226 – ident: ref104 doi: 10.1145/2906388.2906396 – year: 0 ident: ref83 publication-title: VPU – start-page: 1 year: 2017 ident: ref10 article-title: Precog: Prefetching for image recognition applications at the edge publication-title: Proc ACM/IEEE Symp Edge Comput – ident: ref44 doi: 10.1109/ICCV.2015.279 – ident: ref60 doi: 10.1109/TCAD.2018.2858384 – volume: abs 1804 3235 start-page: 1 year: 2018 ident: ref128 article-title: Large scale distributed neural network training through online distillation publication-title: CoRR – year: 0 ident: ref32 publication-title: Cudnn – ident: ref120 doi: 10.1145/3196494.3196522 – ident: ref88 doi: 10.1109/IPSN.2016.7460664 – year: 2015 ident: ref71 article-title: Deep compression: Compressing deep neural networks with pruning, trained quantization and Huffman coding publication-title: arXiv 1510 00149 [cs] – start-page: 29 year: 2018 ident: ref95 article-title: Mainstream: Dynamic stem-sharing for multi-tenant video processing publication-title: Proc Usenix Ann Technical Conf (Usenix '99) – ident: ref53 doi: 10.1109/MNET.2018.1800109 – volume: abs 1409 4842 start-page: 1 year: 2014 ident: ref102 article-title: Going deeper with convolutions publication-title: CoRR – ident: ref141 doi: 10.1145/2462456.2464448 – volume: abs 1508 1991 start-page: 1 year: 2015 ident: ref97 article-title: Bidirectional LSTM-CRF models for sequence tagging publication-title: CoRR – volume: abs 1609 4747 start-page: 1 year: 2016 ident: ref15 article-title: An overview of gradient descent optimization algorithms publication-title: CoRR – ident: ref113 doi: 10.1145/3132211.3134459 – ident: ref76 doi: 10.1109/ICPR.2016.7900006 – ident: ref142 doi: 10.1145/3081333.3081347 – year: 2018 ident: ref135 article-title: Privacy-preserving machine learning through data obfuscation publication-title: arXiv 1807 01860 – year: 0 ident: ref82 publication-title: Accelerating deep convolutional neural networks using specialized hardware – ident: ref74 doi: 10.1145/2994551.2994564 – volume: abs 1709 1686 start-page: 1 year: 2017 ident: ref110 article-title: BranchyNet: Fast inference via early exiting from deep neural networks publication-title: CoRR – ident: ref4 doi: 10.1109/MPRV.2009.82 – ident: ref138 doi: 10.1016/S0893-6080(98)00116-6 – year: 2019 ident: ref3 publication-title: Related's Hudson Yards Smart City or Surveillance City? – year: 0 ident: ref29 publication-title: Caffe – volume: abs 1812 5374 start-page: 1 year: 2018 ident: ref54 article-title: Distributed deep learning at the edge: A novel proactive and cooperative caching framework for mobile edge networks publication-title: CoRR – year: 2015 ident: ref75 article-title: Distilling the knowledge in a neural network publication-title: ArXiv 1503 02531 – start-page: 1310 year: 2015 ident: ref14 article-title: Privacy-preserving deep learning publication-title: Proc 22nd ACM SIGSAC Conf Comput Commun Secur doi: 10.1145/2810103.2813687 – ident: ref62 doi: 10.1145/3229625.3229629 – start-page: 1651 year: 2018 ident: ref119 article-title: GAZELLE: A low latency framework for secure neural network inference publication-title: Proc 27th USENIX Security Symp (USENIX Security) – year: 0 ident: ref40 publication-title: Deep Learning for Siri's Voice On-Device Deep Mixture Density Networks for Hybrid Unit Selection Synthesis – ident: ref61 doi: 10.1109/COMST.2018.2844341 – ident: ref63 doi: 10.1109/CVPR.2018.00559 – ident: ref64 doi: 10.1145/3097895.3097896 – volume: abs 1604 981 start-page: 1 year: 2016 ident: ref123 article-title: Revisiting distributed synchronous SGD publication-title: CoRR – ident: ref57 doi: 10.1109/ICCV.2013.257 – start-page: 629 year: 2017 ident: ref127 article-title: Gaia: Geo-distributed machine learning approaching LAN speeds publication-title: Proc USENIX Symp Netw Syst Design Implem (NSDI) – ident: ref98 doi: 10.1162/neco.1997.9.8.1735 – start-page: 82 year: 2017 ident: ref78 article-title: DeepMon: Mobile GPU-based deep learning framework for continuous vision applications publication-title: Proc ACM MOBISYS – ident: ref106 doi: 10.1109/MNET.2018.1700202 – ident: ref90 doi: 10.1109/MCOM.2018.1701277 – start-page: 91 year: 2015 ident: ref36 article-title: Faster R-Cnn: Towards real-time object detection with region proposal networks publication-title: Proc Adv Neural Inf Process Syst – ident: ref72 doi: 10.1145/3240765.3243473 – ident: ref137 doi: 10.1109/SEC.2018.00014 – start-page: 1097 year: 2012 ident: ref99 article-title: ImageNet classification with deep convolutional neural networks publication-title: Proc Adv Neural Inf Process Syst – ident: ref134 doi: 10.1145/2976749.2978318 – ident: ref41 doi: 10.18653/v1/N16-1030 – volume: abs 1512 3385 start-page: 1 year: 2015 ident: ref103 article-title: Deep residual learning for image recognition publication-title: CoRR – ident: ref89 doi: 10.1109/JPROC.2017.2761740 – ident: ref117 doi: 10.1145/3133956.3134056 – year: 2017 ident: ref21 article-title: MobileNets: Efficient convolutional neural networks for mobile vision applications publication-title: arXiv 1704 04861 – start-page: 201 year: 2016 ident: ref116 article-title: CryptoNets: Applying neural networks to encrypted data with high throughput and accuracy publication-title: Proc Int Conf Mach Learn – ident: ref91 doi: 10.1145/2809695.2809711 – year: 0 ident: ref28 publication-title: Caffe – start-page: 377 year: 2017 ident: ref93 article-title: Live video analytics at scale with approximation and delay-tolerance publication-title: Proc USENIX NSDI – year: 2016 ident: ref132 article-title: How to scale distributed deep learning? publication-title: arXiv 1611 04581 – ident: ref114 doi: 10.1145/1999995.2000000 – year: 0 ident: ref85 publication-title: NVIDIA EGX Edge Computing Platform – start-page: 9017 year: 2018 ident: ref47 article-title: FastGRNN: A fast, accurate, stable and tiny kilobyte sized gated recurrent neural network publication-title: Proc Adv Neural Inf Process Syst – start-page: 12 year: 2017 ident: ref143 article-title: You can teach elephants to dance: Agile VM handoff for edge computing publication-title: Proc Symp Comput Commun – start-page: 265 year: 2006 ident: ref150 article-title: Calibrating noise to sensitivity in private data analysis publication-title: Proc Theory Cryptogr Conf doi: 10.1007/11681878_14 – volume: 70 start-page: 1935 year: 2017 ident: ref59 article-title: Resource-efficient machine learning in 2 KB RAM for the Internet of Things publication-title: Proc 34th Int Conf Mach Learn – volume: abs 1712 8132 start-page: 1 year: 2017 ident: ref55 article-title: A deep reinforcement learning-based framework for content caching publication-title: CoRR – ident: ref2 doi: 10.1109/MC.2017.9 – year: 0 ident: ref30 publication-title: PyTorch – ident: ref50 doi: 10.14722/ndss.2018.23204 – ident: ref147 doi: 10.1109/INFCOM.2012.6195689 – ident: ref115 doi: 10.1145/3219819.3220106 – year: 0 ident: ref48 publication-title: Google's Pixel Buds are no Match for Professional Interpreters – year: 2016 ident: ref42 article-title: Google's neural machine translation system: Bridging the gap between human and machine translation publication-title: arXiv 1609 08144 – volume: abs 1712 1887 start-page: 1 year: 2017 ident: ref129 article-title: Deep gradient compression: Reducing the communication bandwidth for distributed training publication-title: CoRR – year: 2017 ident: ref46 publication-title: Hey Siri An On-Device DNN-Powered Voice Trigger for Apple's Personal Assistant – ident: ref81 doi: 10.1145/2996864 – ident: ref11 doi: 10.1109/MNET.2018.1700202 – year: 2018 ident: ref65 publication-title: Enabling full body AR with Mask R-CNN2Go – ident: ref149 doi: 10.1109/SP.2017.41 – ident: ref136 doi: 10.1109/SP.2017.12 – ident: ref107 doi: 10.23919/DATE.2017.7927211 – ident: ref122 doi: 10.1109/SEC.2016.38 – start-page: 1 year: 2017 ident: ref34 article-title: DeepIoT: Compressing deep neural network structures for sensing systems with a compressor-critic framework publication-title: Proc SenSys – volume: abs 1603 4467 start-page: 1 year: 2016 ident: ref25 article-title: TensorFlow: Large-scale machine learning on heterogeneous distributed systems publication-title: CoRR – ident: ref96 doi: 10.1109/5.726791 – ident: ref20 doi: 10.1109/INFOCOM.2018.8485905 – ident: ref146 doi: 10.1145/2999572.2999602 – ident: ref19 doi: 10.1145/3123266.3123389 – start-page: 685 year: 2015 ident: ref124 article-title: Deep learning with elastic averaging SGD publication-title: Advances in neural information processing systems – year: 0 ident: ref26 publication-title: Tensorflowlite – year: 0 ident: ref45 publication-title: Pagespeed Insights Improve Server Response Time – ident: ref112 doi: 10.1145/1814433.1814441 – year: 0 ident: ref31 publication-title: Cuda
SSID	ssj0003563
Score	2.727748
SecondaryResourceType	review_article
Snippet	Deep learning is currently widely used in a variety of applications, including computer vision and natural language processing. End devices, such as...
SourceID	proquest crossref ieee
SourceType	Aggregation Database Enrichment Source Index Database Publisher
StartPage	1655
SubjectTerms	Artificial intelligence Cloud computing Computational modeling Computer vision Deep learning Devices Edge computing Finite element method Inference Machine learning mobile computing Natural language processing Neural networks Privacy Servers Smartphones State-of-the-art reviews Training
Title	Deep Learning With Edge Computing: A Review
URI	https://ieeexplore.ieee.org/document/8763885 https://www.proquest.com/docview/2270183992
Volume	107
WOSCitedRecordID	wos000497973300011&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText	1
inHoldings	1
isFullTextHit
isPrint
journalDatabaseRights	– providerCode: PRVIEE databaseName: IEEE Electronic Library (IEL) customDbUrl: eissn: 1558-2256 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0003563 issn: 0018-9219 databaseCode: RIE dateStart: 19630101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LS8NAEB7a4kEPvqpYrbIHb5o2r315K6UiIrWIYm9hX6kFaUub-vvdTdKgKIK3kOyEMJt57O7M9wFcEofIwgn3gjRIvVgr35NCaI-z2AZLHitcsJY80OGQjcd8VIPrqhfGGJMXn5mOu8zP8vVcrd1WWdehpzGG61CnlBa9WpXXjXDJmhZYA7ZmuGmQ8Xn3fvT02HdVXLwTukeUfgtCOavKD1ecx5fbvf992T7slnkk6hUTfwA1MzuEnS_ogk2wvsQsUAmgOkGv0-wNDfTEoILJwd67QT1UHA4cwcvt4Ll_55XcCJ4KOc48nTpuPYkjqn2c8tgoEhgZS4w1N4SISCgeR0xgB7EeSxOFwi50sCRC-JrJNDqGxmw-MyeAqCaMCj_WgbRrCy4F036qrBCx6YQmvAXBRlmJKoHDHX_Fe5IvIHye5ApOnIKTUsEtuKpkFgVsxp-jm06l1chSmy1ob-YkKS1rlYQh9V1Wx8PT36XOYNu9uyjSa0MjW67NOWypj2y6Wl7kP80nnLy7qA
linkProvider	IEEE
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LT8JAEJ4gmqgHX2hEUXvwpoU-drddb8RAUBGJwcit2VeRxACB4u93ty2NRmPirWl30ma289jdme8DuCQGkYUSaruxG9tICsfmjEmbhkgHS4oEzlhLukGvFw6HtF-C66IXRimVFp-purlMz_LlVCzNVlnDoKeFIV6DdYyQ52bdWoXf9XHOm-ZqE9aGuGqRcWjjvv_8dGvquGjdM4-C4FsYSnlVfjjjNMK0d__3bXuwk2eSVjOb-n0oqckBbH_BF6yA9iZqZuUQqiPrdZy8WS05UlbG5aDv3VhNKzseOISXdmtw27FzdgRbeBQntowNux7HfiAdHFOkBHEVRxxjSRUhzGeCIj9k2ICsI658j-mlDuaEMUeGPPaPoDyZTtQxWIEkYcAcJF2uVxeUs1A6sdBCRCcUktAquCtlRSKHDjcMFu9RuoRwaJQqODIKjnIFV-GqkJllwBl_jq4YlRYjc21Wobaakyi3rUXkeYFj8jrqnfwudQGbncFjN-re9R5OYcu8JyvZq0E5mS_VGWyIj2S8mJ-nP9Anj1--7w
openUrl	ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Deep+Learning+With+Edge+Computing%3A+A+Review&rft.jtitle=Proceedings+of+the+IEEE&rft.au=Chen%2C+Jiasi&rft.au=Ran%2C+Xukan&rft.date=2019-08-01&rft.pub=IEEE&rft.issn=0018-9219&rft.volume=107&rft.issue=8&rft.spage=1655&rft.epage=1674&rft_id=info:doi/10.1109%2FJPROC.2019.2921977&rft.externalDocID=8763885
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0018-9219&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0018-9219&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0018-9219&client=summon