# Citywide Cellular Traffic Prediction Based on a Hybrid Spatiotemporal Network

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Data Observation and Analysis

#### 2.1. Temporal Domain

#### 2.2. Spatial Domain

## 3. Cellular Traffic Prediction Model

#### 3.1. Model Framework Introduction

#### 3.2. Convolution Module

#### 3.3. Deformable Convolution

#### 3.4. Time Embedding Module

- Dividing the time period of the day into 24 segments, representing 24 h, the time attribute of each data was represented by a 24-dimensional one-hot vector (Hour_of_Day).
- Holiday (including weekends and Italian festivals) is represented by a one-dimensional vector (Is_of_Holiday) and is entered with 0 or 1, 1 indicating that the day is a holiday and 0 indicating that the day is a working day.

^{th}fully connected layer. $\sigma $ represents the sigmoid activation function.

#### 3.5. Attention Module

## 4. Experimental Results and Analysis

#### 4.1. Experimental Process and Parameter Setting

#### 4.2. Experiment Analysis

#### 4.3. Experimental Result

## 5. Conclusions

- This work used DenseNet with deformable convolution to extract the spatiotemporal characteristics of traffic.
- We introduced hour and holiday information to aid traffic forecasting.
- We proposed an attention module based on historical data to adjust the weight of the predicted traffic.

- The model did not have a good ability to respond to fluctuations caused by emergencies.
- The forecast performance of the large scale traffic volume (total traffic volume of the entire city) needs to be improved.
- There are many external factors that we did not consider that could have a potential impact on cellular traffic changes.

## Author Contributions

## Funding

## Acknowledgments

## Conflicts of Interest

## References

- Li, S.; Da Xu, L.; Zhao, S. 5G Internet of Things: A survey. J. Ind. Inf. Integr.
**2018**, 10, 1–9. [Google Scholar] [CrossRef] - Javaid, N.; Sher, A.; Nasir, H.; Guizani, N. Intelligence in IoT-based 5G networks: Opportunities and challenges. IEEE Commun. Mag.
**2018**, 56, 94–100. [Google Scholar] [CrossRef] - Wollschlaeger, M.; Sauter, T.; Jasperneite, J. The future of industrial communication: Automation networks in the era of the internet of things and industry 4.0. IEEE Ind. Electron. Mag.
**2017**, 11, 17–27. [Google Scholar] [CrossRef] - Akpakwu, G.A.; Silva, B.J.; Hancke, G.P.; Abu-Mahfouz, A.M. A survey on 5G networks for the Internet of Things: Communication technologies and challenges. IEEE Access
**2017**, 6, 3619–3647. [Google Scholar] [CrossRef] - Rao, S.K.; Prasad, R. Impact of 5G technologies on industry 4.0. Wirel. Pers. Commun.
**2018**, 100, 145–159. [Google Scholar] [CrossRef] - Huang, Y.; Tan, J.; Liang, Y.C. Wireless big data: Transforming heterogeneous networks to smart networks. J. Commun. Inf. Netw.
**2017**, 2, 19–32. [Google Scholar] [CrossRef] [Green Version] - Bi, S.; Zhang, R.; Ding, Z.; Cui, S. Wireless communications in the era of big data. IEEE Commun. Mag.
**2015**, 53, 190–199. [Google Scholar] [CrossRef] [Green Version] - Yao, C.; Yang, C.; Chih-Lin, I. Data-driven resource allocation with traffic load prediction. J. Commun. Inf. Netw.
**2017**, 2, 52–65. [Google Scholar] [CrossRef] [Green Version] - Shu, Y.; Yu, M.; Yang, O.; Liu, J.; Feng, H. Wireless traffic modeling and prediction using seasonal ARIMA models. IEICE Trans. Commun.
**2005**, 88, 3992–3999. [Google Scholar] [CrossRef] - Zhou, B.; He, D.; Sun, Z. Traffic predictability based on ARIMA/GARCH model. In Proceedings of the 2006 2nd Conference on Next Generation Internet Design and Engineering, Valencia, Spain, 3–5 April 2006. [Google Scholar]
- Li, R.; Zhao, Z.; Zheng, J.; Mei, C.; Cai, Y.; Zhang, H. The learning and prediction of application-level traffic data in cellular networks. IEEE Trans. Wirel. Commun.
**2017**, 16, 3899–3912. [Google Scholar] [CrossRef] - Chen, X.; Jin, Y.; Qiang, S.; Hu, W.; Jiang, K. Analyzing and modeling spatio-temporal dependence of cellular traffic at city scale. In Proceedings of the 2015 IEEE International Conference on Communications (ICC), London, UK, 8–12 June 2015. [Google Scholar]
- Wang, J.; Gong, B.; Liu, H.; Li, S. Multidisciplinary approaches to artificial swarm intelligence for heterogeneous computing and cloud scheduling. Appl. Intell.
**2015**, 43, 662–675. [Google Scholar] [CrossRef] - Sun, H.; Liu, H.X.; Xiao, H.; He, R.R.; Ran, B. Use of local linear regression model for short-term traffic forecasting. Transp. Res. Rec.
**2003**, 1836, 143–150. [Google Scholar] [CrossRef] - Sapankevych, N.I.; Sankar, R. Time series prediction using support vector machines: A survey. IEEE Comput. Intell. Mag.
**2009**, 4, 24–38. [Google Scholar] [CrossRef] - Wang, J.; Tang, J.; Xu, Z.; Wang, Y.; Xue, G.; Zhang, X.; Yang, D. Spatiotemporal modeling and prediction in cellular networks: A big data enabled deep learning approach. In Proceedings of the IEEE INFOCOM 2017—IEEE Conference on Computer Communications, Atlanta, GA, USA, 1–4 May 2017. [Google Scholar]
- Zhang, C.; Zhang, H.; Yuan, D.; Zhang, M. Citywide cellular traffic prediction based on densely connected convolutional neural networks. IEEE Commun. Lett.
**2018**, 22, 1656–1659. [Google Scholar] [CrossRef] - Barlacchi, G.; De Nadai, M.; Larcher, R.; Casella, A.; Chitic, C.; Torrisi, G.; Antonelli, F.; Vespignani, A.; Pentland, A.; Lepri, B. A multi-source dataset of urban life in the city of Milan and the Province of Trentino. Sci. Data
**2015**, 2, 150055. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Huang, G.; Liu, Z.; Van Der Maaten, L.; Weinberger, K.Q. Densely connected convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA, 21–26 July 2017. [Google Scholar]
- Dai, J.; Qi, H.; Xiong, Y.; Li, Y.; Zhang, G.; Hu, H.; Wei, Y. Deformable convolutional networks. In Proceedings of the IEEE International Conference on Computer Vision, Venice, Italy, 22–29 October 2017. [Google Scholar]
- He, K.; Zhang, X.; Ren, S.; Sun, J. Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA, 21–26 July 2017. [Google Scholar]
- Kingma, D.P.; Ba, J. Adam: A method for stochastic optimization. arXiv
**2014**, arXiv:1412.6980. [Google Scholar]

Dataset | Model | MAE | RMSE |
---|---|---|---|

SMS | STDenseNet | 11.10 | 27.49 |

+DeformConv | 10.81 | 26.91 | |

+Time-property | 10.66 | 27.22 | |

+Attention | 10.09 | 26.62 | |

HSTNet | 10.01 | 26.42 | |

Call | STDenseNet | 8.13 | 17.10 |

+DeformConv | 7.61 | 16.18 | |

+Time-property | 8.03 | 16.89 | |

+Attention | 7.27 | 16.70 | |

HSTNet | 7.25 | 16.04 | |

Internet | STDenseNet | 44.15 | 80.51 |

+DeformConv | 43.23 | 77.75 | |

+Time-property | 39.73 | 77.08 | |

+Attention | 39.89 | 74.48 | |

HSTNet | 39.19 | 72.72 |

Model | Time | Parameters |
---|---|---|

STDenseNet | 22s | 239K |

+DeformConv | 34s | 170K |

+Time-property | 23s | 350K |

+Attention | 22s | 243K |

HSTNet | 35s | 284K |

Input Dimension | 1 | 2 | 3 | 4 |
---|---|---|---|---|

SMS | 27.51 | 27.18 | 26.42 | 26.83 |

Call | 16.86 | 16.23 | 16.04 | 16.62 |

Internet | 80.10 | 75.38 | 72.72 | 78.32 |

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Zhang, D.; Liu, L.; Xie, C.; Yang, B.; Liu, Q.
Citywide Cellular Traffic Prediction Based on a Hybrid Spatiotemporal Network. *Algorithms* **2020**, *13*, 20.
https://doi.org/10.3390/a13010020

**AMA Style**

Zhang D, Liu L, Xie C, Yang B, Liu Q.
Citywide Cellular Traffic Prediction Based on a Hybrid Spatiotemporal Network. *Algorithms*. 2020; 13(1):20.
https://doi.org/10.3390/a13010020

**Chicago/Turabian Style**

Zhang, Dehai, Linan Liu, Cheng Xie, Bing Yang, and Qing Liu.
2020. "Citywide Cellular Traffic Prediction Based on a Hybrid Spatiotemporal Network" *Algorithms* 13, no. 1: 20.
https://doi.org/10.3390/a13010020