Abstract
Future communication networks promise to provide ubiquitous high-speed services for numerous users via densely deployed small cells. They should offer good user experiences to all the users while incurring a low operational cost to the operators. User scheduling is a well-known approach to deliver good user experience, and recent works further demonstrate that it is also beneficial to improve energy efficiency (EE). However, existing EE-based scheduling schemes tend to favor users with good channel condition which lead to unfair user experiences. In this paper, we introduce a new concept of resource allocation boundary where EE and user fairness can be addressed simultaneously. We derive the boundary that partition in an effective manner the users into different groups. By applying an appropriate scheduling strategy to each group of users, not only users with poorer channel conditions can be served fairly, but the EE of the system can be further improved. We also provide a low-complexity energy-efficient power allocation algorithm that is designed to fully exploit the transmit power reduction capability of small cells. Simulation results show that our new scheduling scheme can improve the EE and user fairness when compared to existing approaches, i.e. by up to 63% and 56%, respectively.