English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Cellular traffic prediction is of great importance on the path of enabling 5Gmobile networks to perform intelligent and efficient infrastructure planningand management. However, available data are limited to base station logginginformation. Hence, training methods for generating high-quality predictionsthat can generalize to new observations across diverse parties are in demand.Traditional approaches require collecting measurements from multiple basestations, transmitting them to a central entity and conducting machine learningoperations using the acquire data. The dissemination of local observationsraises concerns regarding confidentiality and performance, which impede theapplicability of machine learning techniques. Although various distributedlearning methods have been proposed to address this issue, their application totraffic prediction remains highly unexplored. In this work, we investigate theefficacy of federated learning applied to raw base station LTE data fortime-series forecasting. We evaluate one-step predictions using five differentneural network architectures trained with a federated setting onnon-identically distributed data. Our results show that the learningarchitectures adapted to the federated setting yield equivalent predictionerror to the centralized setting. In addition, preprocessing techniques on basestations enhance forecasting accuracy, while advanced federated aggregators donot surpass simpler approaches. Simulations considering the environmentalimpact suggest that federated learning holds the potential for reducing carbonemissions and energy consumption. Finally, we consider a large-scale scenariowith synthetic data and demonstrate that federated learning reduces thecomputational and communication costs compared to centralized settings.

Federated Learning for 5G Base Station Traffic Forecasting