A Survey of Machine Learning in Edge Computing: Techniques, Frameworks, Applications, Issues, and Research Directions
Internet of Things (IoT) devices often operate with limited resources while interacting with users and their environment, generating a wealth of data. Machine learning models interpret such sensor data, enabling accurate predictions and informed decisions. However, the sheer volume of data from bill...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-06-01
|
| Series: | Technologies |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7080/12/6/81 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841550983666073600 |
|---|---|
| author | Oumayma Jouini Kaouthar Sethom Abdallah Namoun Nasser Aljohani Meshari Huwaytim Alanazi Mohammad N. Alanazi |
| author_facet | Oumayma Jouini Kaouthar Sethom Abdallah Namoun Nasser Aljohani Meshari Huwaytim Alanazi Mohammad N. Alanazi |
| author_sort | Oumayma Jouini |
| collection | DOAJ |
| description | Internet of Things (IoT) devices often operate with limited resources while interacting with users and their environment, generating a wealth of data. Machine learning models interpret such sensor data, enabling accurate predictions and informed decisions. However, the sheer volume of data from billions of devices can overwhelm networks, making traditional cloud data processing inefficient for IoT applications. This paper presents a comprehensive survey of recent advances in models, architectures, hardware, and design requirements for deploying machine learning on low-resource devices at the edge and in cloud networks. Prominent IoT devices tailored to integrate edge intelligence include Raspberry Pi, NVIDIA’s Jetson, Arduino Nano 33 BLE Sense, STM32 Microcontrollers, SparkFun Edge, Google Coral Dev Board, and Beaglebone AI. These devices are boosted with custom AI frameworks, such as TensorFlow Lite, OpenEI, Core ML, Caffe2, and MXNet, to empower ML and DL tasks (e.g., object detection and gesture recognition). Both traditional machine learning (e.g., random forest, logistic regression) and deep learning methods (e.g., ResNet-50, YOLOv4, LSTM) are deployed on devices, distributed edge, and distributed cloud computing. Moreover, we analyzed 1000 recent publications on “ML in IoT” from IEEE Xplore using support vector machine, random forest, and decision tree classifiers to identify emerging topics and application domains. Hot topics included big data, cloud, edge, multimedia, security, privacy, QoS, and activity recognition, while critical domains included industry, healthcare, agriculture, transportation, smart homes and cities, and assisted living. The major challenges hindering the implementation of edge machine learning include encrypting sensitive user data for security and privacy on edge devices, efficiently managing resources of edge nodes through distributed learning architectures, and balancing the energy limitations of edge devices and the energy demands of machine learning. |
| format | Article |
| id | doaj-art-4ae1906889a8471fabf6fec0c1061fdb |
| institution | Kabale University |
| issn | 2227-7080 |
| language | English |
| publishDate | 2024-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Technologies |
| spelling | doaj-art-4ae1906889a8471fabf6fec0c1061fdb2025-01-09T17:40:16ZengMDPI AGTechnologies2227-70802024-06-011268110.3390/technologies12060081A Survey of Machine Learning in Edge Computing: Techniques, Frameworks, Applications, Issues, and Research DirectionsOumayma Jouini0Kaouthar Sethom1Abdallah Namoun2Nasser Aljohani3Meshari Huwaytim Alanazi4Mohammad N. Alanazi5Innov’COM Laboratory, Higher School of Communication of Tunis (SUPCOM), Technopark Elghazala, Ariana 2083, TunisiaInnov’COM Laboratory, Higher School of Communication of Tunis (SUPCOM), Technopark Elghazala, Ariana 2083, TunisiaFaculty of Computer and Information Systems, Islamic University of Madinah, Madinah 42351, Saudi ArabiaFaculty of Computer and Information Systems, Islamic University of Madinah, Madinah 42351, Saudi ArabiaComputer Science Department, College of Sciences, Northern Border University, Arar 91431, Saudi ArabiaCollege of Computer and Information Sciences, Imam Mohammad Ibn Saud Islamic University (IMSIU), Riyadh 13318, Saudi ArabiaInternet of Things (IoT) devices often operate with limited resources while interacting with users and their environment, generating a wealth of data. Machine learning models interpret such sensor data, enabling accurate predictions and informed decisions. However, the sheer volume of data from billions of devices can overwhelm networks, making traditional cloud data processing inefficient for IoT applications. This paper presents a comprehensive survey of recent advances in models, architectures, hardware, and design requirements for deploying machine learning on low-resource devices at the edge and in cloud networks. Prominent IoT devices tailored to integrate edge intelligence include Raspberry Pi, NVIDIA’s Jetson, Arduino Nano 33 BLE Sense, STM32 Microcontrollers, SparkFun Edge, Google Coral Dev Board, and Beaglebone AI. These devices are boosted with custom AI frameworks, such as TensorFlow Lite, OpenEI, Core ML, Caffe2, and MXNet, to empower ML and DL tasks (e.g., object detection and gesture recognition). Both traditional machine learning (e.g., random forest, logistic regression) and deep learning methods (e.g., ResNet-50, YOLOv4, LSTM) are deployed on devices, distributed edge, and distributed cloud computing. Moreover, we analyzed 1000 recent publications on “ML in IoT” from IEEE Xplore using support vector machine, random forest, and decision tree classifiers to identify emerging topics and application domains. Hot topics included big data, cloud, edge, multimedia, security, privacy, QoS, and activity recognition, while critical domains included industry, healthcare, agriculture, transportation, smart homes and cities, and assisted living. The major challenges hindering the implementation of edge machine learning include encrypting sensitive user data for security and privacy on edge devices, efficiently managing resources of edge nodes through distributed learning architectures, and balancing the energy limitations of edge devices and the energy demands of machine learning.https://www.mdpi.com/2227-7080/12/6/81machine learningInternet of ThingsIoT devicesedge intelligenceedge learningartificial intelligence |
| spellingShingle | Oumayma Jouini Kaouthar Sethom Abdallah Namoun Nasser Aljohani Meshari Huwaytim Alanazi Mohammad N. Alanazi A Survey of Machine Learning in Edge Computing: Techniques, Frameworks, Applications, Issues, and Research Directions Technologies machine learning Internet of Things IoT devices edge intelligence edge learning artificial intelligence |
| title | A Survey of Machine Learning in Edge Computing: Techniques, Frameworks, Applications, Issues, and Research Directions |
| title_full | A Survey of Machine Learning in Edge Computing: Techniques, Frameworks, Applications, Issues, and Research Directions |
| title_fullStr | A Survey of Machine Learning in Edge Computing: Techniques, Frameworks, Applications, Issues, and Research Directions |
| title_full_unstemmed | A Survey of Machine Learning in Edge Computing: Techniques, Frameworks, Applications, Issues, and Research Directions |
| title_short | A Survey of Machine Learning in Edge Computing: Techniques, Frameworks, Applications, Issues, and Research Directions |
| title_sort | survey of machine learning in edge computing techniques frameworks applications issues and research directions |
| topic | machine learning Internet of Things IoT devices edge intelligence edge learning artificial intelligence |
| url | https://www.mdpi.com/2227-7080/12/6/81 |
| work_keys_str_mv | AT oumaymajouini asurveyofmachinelearninginedgecomputingtechniquesframeworksapplicationsissuesandresearchdirections AT kaoutharsethom asurveyofmachinelearninginedgecomputingtechniquesframeworksapplicationsissuesandresearchdirections AT abdallahnamoun asurveyofmachinelearninginedgecomputingtechniquesframeworksapplicationsissuesandresearchdirections AT nasseraljohani asurveyofmachinelearninginedgecomputingtechniquesframeworksapplicationsissuesandresearchdirections AT mesharihuwaytimalanazi asurveyofmachinelearninginedgecomputingtechniquesframeworksapplicationsissuesandresearchdirections AT mohammadnalanazi asurveyofmachinelearninginedgecomputingtechniquesframeworksapplicationsissuesandresearchdirections AT oumaymajouini surveyofmachinelearninginedgecomputingtechniquesframeworksapplicationsissuesandresearchdirections AT kaoutharsethom surveyofmachinelearninginedgecomputingtechniquesframeworksapplicationsissuesandresearchdirections AT abdallahnamoun surveyofmachinelearninginedgecomputingtechniquesframeworksapplicationsissuesandresearchdirections AT nasseraljohani surveyofmachinelearninginedgecomputingtechniquesframeworksapplicationsissuesandresearchdirections AT mesharihuwaytimalanazi surveyofmachinelearninginedgecomputingtechniquesframeworksapplicationsissuesandresearchdirections AT mohammadnalanazi surveyofmachinelearninginedgecomputingtechniquesframeworksapplicationsissuesandresearchdirections |