Traffic Load Prediction and Power Consumption Reduction for Multi-band Networks

Energy is a major expense issue for mobile operators. In the case of wireless networks, base stations have been identified as the main source of energy consumption. In this paper, we study the energy consumption reduction problem based on real measurements for a commercial multi-band LTE network. Sp...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Diouf, Ndolane, Anamuro, Cesar Vargas, Gueguen, Cédric, Ndong, Massa, Talla, Kharouna, Lagrange, Xavier
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Energy is a major expense issue for mobile operators. In the case of wireless networks, base stations have been identified as the main source of energy consumption. In this paper, we study the energy consumption reduction problem based on real measurements for a commercial multi-band LTE network. Specifically, we are interested in sleep modes to turn off certain frequency bands during low traffic periods and consequently reduce power consumption. We determine the number of frequency bands really needed at each time period. The frequency bands that are not needed can be disabled to reduce energy consumption. In order to allow the operator to predict how many bands can be switched off without major impact on the quality of service, we propose to use a deep learning algorithm, such as Long-Short Term Memory (LSTM). Based on the captured data traces, we have shown that the proposed LSTM model can save an average of 8% to 21% of the energy consumption during working days.
DOI:10.48550/arxiv.2312.11958