Conference Publication Details
Mandatory Fields
Carolan E.;McLoone S.;Farrell R.
7th EAI International Conference on Mobile Networks and Management
A predictive model for minimising power usage in radio access networks
0 ()
Optional Fields
Cellular networks Cellular usage Green networks Spectrum sharing Temporal dynamics Traffic prediction
© Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2015. In radio access networks traffic load varies greatly both spatially and temporally. However, resource usage of Base Stations (BSs) does not solely depend on the traffic load; auxiliary devices contribute to resource usage in a load invariant manner. Consequently, BSs suffer from a large underutilisation of resources throughout most of the day due to their optimisation for peak traffic hours. In this paper an energy saving scheme is proposed with the use of an Artificial Neural Network (ANN) predictive model to make switching decisions ahead of time. The optimum set of BS to turn off while maintaining Quality Of Service (QoS) is formulated as a binary integer programming problem. We validated our model and found large potential savings using an extensive data set spanning all network usage for three months and over one thousand BSs covering the entirety of Dublin city and county.
Grant Details