In the ultra-dense network scenario

a power allocation scheme based on the utility function optimization model is proposed to maximize the sum rate of user-centric overlapped virtual cell systems. The optimal power allocation is obtained based on the hierarchical alternating iterative algorithm. Firstly

the original problem is decomposed into two sub-problems: power splitting and virtual power allocation. Secondly

an optimal power splitting algorithm based on equilibrium strategy is obtained by introducing the Lagrangian function at the upper layer

and further power allocation is achieved by using the virtual power at the bottom layer. Since the power splitting coefficient takes the form of eigenvector which maximizing Signal to Leakage plus Noise Ratio（SLNR）

a balanced effect on signal enhancement and interference reduction is achieved. Simulation results show that the proposed algorithm is superior to conventional power allocation algorithms in performance

which can effectively reduce the interference and increase the sum rate of the system.