Details

Autor(en) / Beteiligte

• Lin, Gui-Dong
• Xu, Jun-Hua
• Tan, Jing-Yu
• Xue, Ji-Zhong
• Wang, Yu

Titel

A Cross-Time Zone Transfer Consumption Model for Base Station Computing Tasks Based on Photovoltaic Prediction

Ist Teil von

• 2023 International Conference on Machine Learning and Cybernetics (ICMLC), 2023, p.558-563

Ort / Verlag

IEEE

Erscheinungsjahr

2023

Quelle

IEEE Xplore

Beschreibungen/Notizen

• The rapid development of artificial intelligence is inseparable from 5G technology. However, the energy consumption of 5G base station also affects the construction of 5G to a great extent. In recent years, 5G base stations have also further reduced power consumption by combining with new energy consumption technology. However, the uncertainty of photovoltaic power generation also brings trouble to the energy planning of communication operators. This paper proposes a cross-time zone transfer consumption model for base station group computing tasks based on photovoltaic prediction. The photovoltaic prediction model based on LSTM is used to predict the photo-voltaic power. Aiming at the problem of different photovoltaic power generation across time zones, energy planning is carried out according to the energy consumption of local BBU. Through experimental simulation, it is verified that the model can reduce the power cost of communication operators and improve the consumption of new energy.