Non-Cooperative Game-Based Power Allocation for Energy-Efficient NOMA Heterogeneous Network

This paper focus on the two-tier downlink heterogeneous networks (HetNets) with non-orthogonal multiple access (NOMA). The aim is to jointly optimize user scheduling and power allocation to solve the problem of maximizing the energy efficiency (EE) of the NOMA HetNets. We formulate the basic principle of NOMA and develop the EE model of joint user scheduling, optimal power allocation factor among users on the same subchannel and power allocation across subchannels. Firstly, considering that the problem of formulation is a multi-objective optimization (MOO) problem, which is non-convex and NP-hard, we decouple the MOO problem into two single-objective optimization (SOO) problems: 1) joint the optimal power allocation factor of maximizing the EE among users on the same subchannel to solves the subchannel matching, and 2) the power allocation across subchannels. For subchannel matching, non-cooperative game and global optimal search (GOS) are respectively used to solve the power allocation factor among users on the same subchannel and user scheduling. For the power allocation factor among users on the same subchannel, the existence of Nash equilibrium (NE) is proved by introducing super-modular game. Besides, the Nash equilibrium point (NEP) of the game can be obtained by using the proposed algorithm. Then, for the correlation between users-subchannels, this paper proposes a non-cooperative game based user-subchannel global optimal search matching algorithm (NCG-US-GOSMA) with low-complexity. Finally, according to the obtained subchannel matching, the non-convex problem of power allocation across subchannels is converted into convex problem by successive convex approximation (SCA), and then solve it by iteration. Simulation results verify the effectiveness of the proposed scheme; in addition, the EE performance is better than the existing resource allocation algorithm and OFDMA scheme.