Energy-saving Resource Allocation by Exploiting the Context Information

TL;DR

This paper proposes a novel energy-efficient resource allocation policy for wireless base stations that leverages multi-level context information to optimize sleeping, scheduling, and power allocation, significantly reducing energy consumption.

Contribution

It introduces a context-aware resource allocation framework that estimates future traffic using multiple context levels, improving energy efficiency over traditional methods.

Findings

Significant energy savings demonstrated in simulations.

Different context levels contribute uniquely to energy and outage reduction.

Policy outperforms non-context-aware approaches.

Abstract

Improving energy efficiency of wireless systems by exploiting the context information has received attention recently as the smart phone market keeps expanding. In this paper, we devise energy-saving resource allocation policy for multiple base stations serving non-real-time traffic by exploiting three levels of context information, where the background traffic is assumed to occupy partial resources. Based on the solution from a total energy minimization problem with perfect future information,a context-aware BS sleeping, scheduling and power allocation policy is proposed by estimating the required future information with three levels of context information. Simulation results show that our policy provides significant gains over those without exploiting any context information. Moreover, it is seen that different levels of context information play different roles in saving energy and reducing outage in transmission.