Net Utility-Oriented Mobility Load Balancing in Self-Optimizing Networks

Mobility load balancing (MLB) is an important use case in self-optimizing networks. Ping-pong effect is significantly challenging the current design of MLB scheme with low convergence rate, which is highly caused by the hidden-cell phenomenon. In this work, a net utility-oriented MLB scheme is proposed with the hidden-cell effect into consideration, which is triggered among multiple involved cells. In detail, the proposed scheme is with the following characteristics: to guarantee the predicable and controllable fast convergence rate: we design a traffic load margin-aware off-loading threshold; we introduce a refinement step of load transferring; to handle the hidden-cell effect; and a net utility function is presented to reflect the preference characteristics during MLB. The detailed rationale and mathematical modeling of the margin-aware off-loading threshold, the refinement step of load transferring and the net utility function are introduced one by one, based on which we propose a net utility-oriented MLB scheme. Numerical results show the convergence and performance of the proposed MLB scheme.