Joint User Association and BS Operation for Green Communications in Ultra-Dense Heterogeneous Networks

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.