Education, Science, Technology, Innovation and Life
Open Access
Sign In

Mobile communication base station traffic forecast

Download as PDF

DOI: 10.23977/cpcs.2021.51009 | Downloads: 20 | Views: 1075

Author(s)

Dongxiao Liu 1, Wei Li 1

Affiliation(s)

1 North China University of Technology, Beijing, 100043

Corresponding Author

Dongxiao Liu

ABSTRACT

This paper studies the prediction problem based on the historical time series data of the base station. Established LSTM and ARIMA models to train and predict traffic data respectively. This paper choose to put the model species into each time series, and put the forecast results into the data table according to the best. At the same time, we found that the traffic data of the cell showed three patterns of rising, steady, and falling over time, instead of rising all the time. Our LSTM model uses the ‘elu’ function as the activation function, the number of training rounds is 50, the number of neurons is 1000, and the observation and output sequence is 120, which is closer to the true value in short-term prediction. The ARIMA model retains the trend of the original model and has achieved better results in long-term forecasting.

KEYWORDS

LSTM model, ARIMA model, long-term forecasting

CITE THIS PAPER

Dongxiao Liu, Wei Li. Mobile communication base station traffic forecast. Computing, Performance and Communication Systems (2021) Vol. 5: 52-55. DOI: http://dx.doi.org/10.23977/cpcs.2021.51009

REFERENCES

[1] Paparrizos J, Gravano L. k-shape: Effcient and accurate clustering of time series [C]// Proceedings of the 2015 ACM SIGMOD International Conference on
[2] Gullo F, Ponti G Tagarelli A, et al. A time series representation model for accurate and fast similarity detection [J]. Pattern Recognition. 2009, 42(11): 2998-3014.
[3] Dickey, D.A. and W. A. Fuller (1979), “Distribution of the Estimators for Autoregressive Time Series with a Unit Root, ”Journal of the American Statistical Association, 74, p 427–431
[4] Luo, Yonghong, et al. "Multivariate time series imputation with generative adversarial networks."Advances in Neural Information Processing Systems. 2018.

Downloads: 2024
Visits: 100149

Sponsors, Associates, and Links


All published work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright © 2016 - 2031 Clausius Scientific Press Inc. All Rights Reserved.