| Title |
Big data analysis of the Internet of Things in the digital twins of smart city based on deep learning |
| ID_Doc |
36700 |
| Authors |
Li, XM; Liu, H; Wang, WX; Zheng, Y; Lv, H; Lv, Z |
| Title |
Big data analysis of the Internet of Things in the digital twins of smart city based on deep learning |
| Year |
2022 |
| Published |
|
| DOI |
10.1016/j.future.2021.10.006 |
| Abstract |
The study aims to conduct big data analysis (BDA) on the massive data generated in the smart city Internet of things (IoT), make the smart city change to the direction of fine governance and efficient and safe data processing. Aiming at the multi-source data collected in the smart city, the study introduces the deep learning (DL) algorithm while using BDA, and puts forward the distributed parallelism strategy of convolutional neural network (CNN). Meantime, the digital twins (DTs) and multi-hop transmission technology are introduced to construct the smart city DTs multi-hop transmission IoTBDA system based on DL, and further simulate and analyze the performance of the system. The results reveal that in the energy efficiency analysis of model data transmission, the energy efficiency first increases and then decrease as the minimum energy collected alpha(0) increases. But a more suitable power diversion factor rho is crucial to the signal transmission energy efficiency of the IoT-BDA system. The prediction accuracy of the model is analyzed and it suggests that the accuracy of the constructed system reaches 97.80%, which is at least 2.24% higher than the DL algorithm adopted by other scholars. Regarding the data transmission performance of the constructed system, it is found that when the successful transmission probability is 100% and the exponential distribution parameters lambda is valued 0.01 similar to 0.05, it is the closest to the actual result, and the data delay is the smallest, which is maintained at the ms level. To sum up, improving the smart city's IoT-BDA system using the DL approach can reduce data transmission delay, improve data forecasting accuracy, and offer actual efficacy, providing experimental references for the digital development of smart cities in the future. (C) 2021 Elsevier B.V. All rights reserved. |
| Author Keywords |
Deep learning; Smart city; Digital twins; Internet of Things; Big data analysis |
| Index Keywords |
Index Keywords |
| Document Type |
Other |
| Open Access |
Open Access |
| Source |
Science Citation Index Expanded (SCI-EXPANDED) |
| EID |
WOS:000717744500002 |
| WoS Category |
Computer Science, Theory & Methods |
| Research Area |
Computer Science |
| PDF |
|