IoT Fog Computing Optimization Method Based on Improved Convolutional Neural Network
The rapid development of communication technology has promoted the development of the Internet of Things technology. It has resulted in a scarcity of computing resources for the Internet of Things devices, and limited the further development of the Internet of Things. In order to improve the utiliza...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2024-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/10376066/ |
_version_ | 1797347309906821120 |
---|---|
author | Bing Jing Huimin Xue |
author_facet | Bing Jing Huimin Xue |
author_sort | Bing Jing |
collection | DOAJ |
description | The rapid development of communication technology has promoted the development of the Internet of Things technology. It has resulted in a scarcity of computing resources for the Internet of Things devices, and limited the further development of the Internet of Things. In order to improve the utilization efficiency of the system resources for the Internet of Things devices and promote the further development of the Internet of Things, the continuous Markov decision process model is constructed. The value function approximation algorithm of the convolutional neural network is used to solve the problem. Continuous Markov decision process model is an excellent single-user decision process model, but not optimal for multi-user systems. Using convolutional neural network to solve the value function of continuous Markov decision process model, so that it can be applied to multi-user system. The results show that the average algorithm has growth rates of 0.48 and 0.84, respectively, in comparison to the other two algorithms. The average arrival rate has the least effect on the average delay of the value function approximation algorithm and the greatest influence on its power consumption. With the average arrival rate, the average delay of the algorithm increased by 0.25S and the power consumption by 0.27W. The effectiveness of the value function approximation algorithm based on convolutional neural network surpasses that of the multi-user multi-task offloading algorithm and the queue-aware algorithm, thus applying continuous Markov decision process models to multi-user systems. The study combines the continuous Markov decision process model with the resource decision of IOT devices, resulting in optimized resource scheduling decisions and improved utilization efficiency of IOT devices. |
first_indexed | 2024-03-08T11:44:54Z |
format | Article |
id | doaj.art-d81b8c12b3f84b898dd5980249f69b62 |
institution | Directory Open Access Journal |
issn | 2169-3536 |
language | English |
last_indexed | 2024-03-08T11:44:54Z |
publishDate | 2024-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Access |
spelling | doaj.art-d81b8c12b3f84b898dd5980249f69b622024-01-25T00:00:17ZengIEEEIEEE Access2169-35362024-01-01122398240810.1109/ACCESS.2023.334813310376066IoT Fog Computing Optimization Method Based on Improved Convolutional Neural NetworkBing Jing0https://orcid.org/0009-0006-4411-8432Huimin Xue1https://orcid.org/0009-0005-3652-2809Department of Internet of Things Technology, Shanxi Vocational and Technical College of Finance and Trade, Taiyuan, ChinaDepartment of Internet of Things Technology, Shanxi Vocational and Technical College of Finance and Trade, Taiyuan, ChinaThe rapid development of communication technology has promoted the development of the Internet of Things technology. It has resulted in a scarcity of computing resources for the Internet of Things devices, and limited the further development of the Internet of Things. In order to improve the utilization efficiency of the system resources for the Internet of Things devices and promote the further development of the Internet of Things, the continuous Markov decision process model is constructed. The value function approximation algorithm of the convolutional neural network is used to solve the problem. Continuous Markov decision process model is an excellent single-user decision process model, but not optimal for multi-user systems. Using convolutional neural network to solve the value function of continuous Markov decision process model, so that it can be applied to multi-user system. The results show that the average algorithm has growth rates of 0.48 and 0.84, respectively, in comparison to the other two algorithms. The average arrival rate has the least effect on the average delay of the value function approximation algorithm and the greatest influence on its power consumption. With the average arrival rate, the average delay of the algorithm increased by 0.25S and the power consumption by 0.27W. The effectiveness of the value function approximation algorithm based on convolutional neural network surpasses that of the multi-user multi-task offloading algorithm and the queue-aware algorithm, thus applying continuous Markov decision process models to multi-user systems. The study combines the continuous Markov decision process model with the resource decision of IOT devices, resulting in optimized resource scheduling decisions and improved utilization efficiency of IOT devices.https://ieeexplore.ieee.org/document/10376066/MDPCNNvalue functionIoTfog computing |
spellingShingle | Bing Jing Huimin Xue IoT Fog Computing Optimization Method Based on Improved Convolutional Neural Network IEEE Access MDP CNN value function IoT fog computing |
title | IoT Fog Computing Optimization Method Based on Improved Convolutional Neural Network |
title_full | IoT Fog Computing Optimization Method Based on Improved Convolutional Neural Network |
title_fullStr | IoT Fog Computing Optimization Method Based on Improved Convolutional Neural Network |
title_full_unstemmed | IoT Fog Computing Optimization Method Based on Improved Convolutional Neural Network |
title_short | IoT Fog Computing Optimization Method Based on Improved Convolutional Neural Network |
title_sort | iot fog computing optimization method based on improved convolutional neural network |
topic | MDP CNN value function IoT fog computing |
url | https://ieeexplore.ieee.org/document/10376066/ |
work_keys_str_mv | AT bingjing iotfogcomputingoptimizationmethodbasedonimprovedconvolutionalneuralnetwork AT huiminxue iotfogcomputingoptimizationmethodbasedonimprovedconvolutionalneuralnetwork |