Key points are not available for this paper at this time.
Ultra-dense heterogeneous networks (UDHNs) have been widely regarded as a long-term and reliable solution for enhancing system performance and reducing cost efficiency. To guarantee these gains in UDHNs, some problems need to be properly tackled, which include more capital/operational expenditures, more complicated, severe and average interferences, and more energy consumption caused by network densification. To this end, in this article, the BS (base station) on/off operation is introduced into the design of user association mechanisms in UDHNs with a general multi-slope channel model. Specifically, we jointly perform the BS (on/off) operation and user association to maximize the sum of weighted downlink long-term rates. Such rates are closely related to users' signal strengths and the load levels of associated BSs. Considering the finally formulated problem is in a mixed-integer and fractional form, we first need to convert it into a parameterized polynomial one. Then, we design a three-layer iterative algorithm to find the solution to a transformed problem, where the first layer tries to find the introduced Lagrangian multipliers and parameters using a Newton-like method, the second layer optimizes the user association and on/off indices alternately, and the third layer optimizes these indices separately. Simulation results show that the designed algorithm can achieve better user association performance than other existing ones.
Zhou et al. (Wed,) studied this question.