<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>Due to the limited computing power of the perception layer of the Internet of Things (IoT), the ability to analyse and process the collected complex object information data is insufficient, and it is also difficult to complete the storage of a large amount of collected data. Through convolutional neural network-simple recurrent unit (CNN-SRU) deep learning, we preprocess a large amount of complex data in the perception layer. The data collected by the perception layer are first transmitted to the CNN for simple category screening and analysis, and then they reach the SRU link, which is updated and optimised again, to improve the integrity and accuracy of IoT information collection. The results show that the accuracy of gated recurrent unit (GRU), long–short-term memory (LSTM) and SRU algorithms shows a downward trend under the three error evaluation standards of root mean squared error (RMSE), mean absolute error (MAE) and relative error (RE), from 0.034 to 0.015, 0.028 to 0.012 and 0.024 to 0.013, respectively; in terms of training time. The SRU algorithm is increased by 54.52%; the maximum SRU in terms of data storage is increased to 33.22%; and the maximum SRU reduction in data mining energy consumption is 11.45%. This meets the requirements of IoT applications in big data mining.</p>
</abstract>

Due to the limited computing power of the perception layer of the Internet of Things (IoT), the ability to analyse and process the collected complex object information data is insufficient, and it is also difficult to complete the storage of a large amount of collected data. Through convolutional neural network-simple recurrent unit (CNN-SRU) deep learning, we preprocess a large amount of complex data in the perception layer. The data collected by the perception layer are first transmitted to the CNN for simple category screening and analysis, and then they reach the SRU link, which is updated and optimised again, to improve the integrity and accuracy of IoT information collection. The results show that the accuracy of gated recurrent unit (GRU), long–short-term memory (LSTM) and SRU algorithms shows a downward trend under the three error evaluation standards of root mean squared error (RMSE), mean absolute error (MAE) and relative error (RE), from 0.034 to 0.015, 0.028 to 0.012 and 0.024 to 0.013, respectively; in terms of training time. The SRU algorithm is increased by 54.52%; the maximum SRU in terms of data storage is increased to 33.22%; and the maximum SRU reduction in data mining energy consumption is 11.45%. This meets the requirements of IoT applications in big data mining.

In-depth basic data detection device based on Internet of Things technology

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

{"article-title":"In-depth basic data detection device based on Internet of Things technology"}

Due to the limited computing power of the perception layer of the Internet of Things (IoT), the ability to analyse and process the collected complex object...

In-depth basic data detection device based on Internet of Things technology

Data publikacji: 31 lip 2023

Zakres stron: -

Otrzymano: 26 maj 2022

Przyjęty: 17 sie 2022

DOI: https://doi.org/10.2478/amns.2021.2.00257

Słowa kluczowe
Internet of Things, perception layer, data mining, SRU, CNN

© 2024 Shanyi Xie et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

In-depth basic data detection device based on Internet of Things technology

Data publikacji: 31 lip 2023

Zakres stron: -

Otrzymano: 26 maj 2022

Przyjęty: 17 sie 2022

DOI: https://doi.org/10.2478/amns.2021.2.00257

Słowa kluczoweInternet of Things, perception layer, data mining, SRU, CNN

© 2024 Shanyi Xie et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Słowa kluczowe
Internet of Things, perception layer, data mining, SRU, CNN