An Energy-Efficient User Association Scheme Based on Robust Optimization in Ultra-Dense Networks

In ultra-dense networks, due to the diversity of content requested by users, the traffic demand of each user is actually uncertain. User association schemes based on exact traffic demand seem to be outdated. With economic and operational considerations and environmental issues, energy efficiency has now become a key pillar of communication network design. Therefore, we propose an energy-efficient user association scheme based on robust optimization, which improves the stability of association and the energy efficiency of the system. First, a robust optimization model for user association scheme is designed in ultra-dense network deployments. In this model, we construct an uncertainty set of bandwidth to describe the uncertainty of the user traffic demand. Then we solve the association matrix through an iterative user association algorithm to maximize system energy efficiency. The simulation results show that the proposed user association scheme has better energy efficiency performance than the signal to interference plus noise ratio (SINR)-based user association scheme.