Abstract
The cellular networks in 5G are tending to be heterogeneous and dense small-cell networks (DSNs). The rapid growth of mobile data services has brought a lot of problems like scarcity of spectrum resources, overload of some base stations, low coverage ratio and poor quality of experience. In order to solve these problems, DSNs have been proposed and become a hot topic, and user association plays a pivotal role in enhancing the load balancing, the spectrum efficiency, and the energy efficiency of DSNs. In this paper, we establish two-layer DSNs containing macrocell and picocell. Then we define the load efficiency function as the ratio of the actual user-required rate and the achievable rate of the Base Stations (BSs). In the objective function, we also consider the power penalty factor to prevent part of the BSs increasing their transmit power without limit. Considering the simplicity and tractability, we adopt Lagrange duality theory to find optimal user association solution and provide a low-complexity distributed algorithm that converges to a near-optimal solution with a theoretical performance guarantee. Compared with maximum transmission rate algorithm and maximum energy efficiency algorithm, the simulation results show that our proposed user association algorithm of maximum load efficiency with power penalty factor can serve more users, get higher overall load efficiency and lower outage probability, and the proposed algorithm is simple and tractable.