Network Utility Maximization With Asymmetric User Satisfaction

This letter revisits a network utility maximization considering asymmetric user satisfaction (AUS), i.e., a user is more sensitive to a quality of service (QoS) decrement than an increment. Reflecting AUS, we propose a novel fairness criterion called asymmetrically proportional fairness (APF). To achieve APF resource allocation, we devise a utility function and design a bandwidth allocation algorithm. Analytic and numerical results verify that our algorithm achieves APF equilibrium.