Towards QoS-Based Embedded Machine Learning
Due to various breakthroughs and advancements in machine learning and computer architectures, machine learning models are beginning to proliferate through embedded platforms. Some of these machine learning models cover a range of applications including computer vision, speech recognition, healthcare...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-10-01
|
| Series: | Electronics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2079-9292/11/19/3204 |
| _version_ | 1797479715469000704 |
|---|---|
| author | Tom Springer Erik Linstead Peiyi Zhao Chelsea Parlett-Pelleriti |
| author_facet | Tom Springer Erik Linstead Peiyi Zhao Chelsea Parlett-Pelleriti |
| author_sort | Tom Springer |
| collection | DOAJ |
| description | Due to various breakthroughs and advancements in machine learning and computer architectures, machine learning models are beginning to proliferate through embedded platforms. Some of these machine learning models cover a range of applications including computer vision, speech recognition, healthcare efficiency, industrial IoT, robotics and many more. However, there is a critical limitation in implementing ML algorithms efficiently on embedded platforms: the computational and memory expense of many machine learning models can make them unsuitable in resource-constrained environments. Therefore, to efficiently implement these memory-intensive and computationally expensive algorithms in an embedded computing environment, innovative resource management techniques are required at the hardware, software and system levels. To this end, we present a novel quality-of-service based resource allocation scheme that uses feedback control to adjust compute resources dynamically to cope with the varying and unpredictable workloads of ML applications while still maintaining an acceptable level of service to the user. To evaluate the feasibility of our approach we implemented a feedback control scheduling simulator that was used to analyze our framework under various simulated workloads. We also implemented our framework as a Linux kernel module running on a virtual machine as well as a Raspberry Pi 4 single board computer. Results illustrate that our approach was able to maintain a sufficient level of service without overloading the processor as well as providing an energy savings of almost 20% as compared to the native resource management in Linux. |
| first_indexed | 2024-03-09T21:50:48Z |
| format | Article |
| id | doaj.art-569bca846ba74c5abe3fbd2ea7c02916 |
| institution | Directory Open Access Journal |
| issn | 2079-9292 |
| language | English |
| last_indexed | 2024-03-09T21:50:48Z |
| publishDate | 2022-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Electronics |
| spelling | doaj.art-569bca846ba74c5abe3fbd2ea7c029162023-11-23T20:08:04ZengMDPI AGElectronics2079-92922022-10-011119320410.3390/electronics11193204Towards QoS-Based Embedded Machine LearningTom Springer0Erik Linstead1Peiyi Zhao2Chelsea Parlett-Pelleriti3Fowler School of Engineering, Chapman University, Orange, CA 92866, USAFowler School of Engineering, Chapman University, Orange, CA 92866, USAFowler School of Engineering, Chapman University, Orange, CA 92866, USAFowler School of Engineering, Chapman University, Orange, CA 92866, USADue to various breakthroughs and advancements in machine learning and computer architectures, machine learning models are beginning to proliferate through embedded platforms. Some of these machine learning models cover a range of applications including computer vision, speech recognition, healthcare efficiency, industrial IoT, robotics and many more. However, there is a critical limitation in implementing ML algorithms efficiently on embedded platforms: the computational and memory expense of many machine learning models can make them unsuitable in resource-constrained environments. Therefore, to efficiently implement these memory-intensive and computationally expensive algorithms in an embedded computing environment, innovative resource management techniques are required at the hardware, software and system levels. To this end, we present a novel quality-of-service based resource allocation scheme that uses feedback control to adjust compute resources dynamically to cope with the varying and unpredictable workloads of ML applications while still maintaining an acceptable level of service to the user. To evaluate the feasibility of our approach we implemented a feedback control scheduling simulator that was used to analyze our framework under various simulated workloads. We also implemented our framework as a Linux kernel module running on a virtual machine as well as a Raspberry Pi 4 single board computer. Results illustrate that our approach was able to maintain a sufficient level of service without overloading the processor as well as providing an energy savings of almost 20% as compared to the native resource management in Linux.https://www.mdpi.com/2079-9292/11/19/3204embedded machine-learningedge intelligenceruntime resource management and allocationquality-of-servicefeedback control |
| spellingShingle | Tom Springer Erik Linstead Peiyi Zhao Chelsea Parlett-Pelleriti Towards QoS-Based Embedded Machine Learning Electronics embedded machine-learning edge intelligence runtime resource management and allocation quality-of-service feedback control |
| title | Towards QoS-Based Embedded Machine Learning |
| title_full | Towards QoS-Based Embedded Machine Learning |
| title_fullStr | Towards QoS-Based Embedded Machine Learning |
| title_full_unstemmed | Towards QoS-Based Embedded Machine Learning |
| title_short | Towards QoS-Based Embedded Machine Learning |
| title_sort | towards qos based embedded machine learning |
| topic | embedded machine-learning edge intelligence runtime resource management and allocation quality-of-service feedback control |
| url | https://www.mdpi.com/2079-9292/11/19/3204 |
| work_keys_str_mv | AT tomspringer towardsqosbasedembeddedmachinelearning AT eriklinstead towardsqosbasedembeddedmachinelearning AT peiyizhao towardsqosbasedembeddedmachinelearning AT chelseaparlettpelleriti towardsqosbasedembeddedmachinelearning |