A genetic algorithm based cell switch-off scheme for energy saving in dense cell deployments

The energy consumption of mobile networks is rapidly growing. Operators have both economic and environmental incentives to increase the energy efficiency of their networks. One way of saving energy is to switch off cells during periods of light traffic. However, cell switch-off is a difficult problem to solve through conventional optimization; existing research makes various assumptions to simplify the problem and offers some heuristics to solve it. The problem can be constructed in different ways depending on the system model that is chosen. We examine the cell switch-off problem with the assumption that each user terminal (UT) has a minimum rate requirement, and show that it can be formulated and solved as a binary integer linear programming (BILP) problem when interference is considered to be constant. This formulation is equivalent to the bin-packing problem, which is NP-hard, if the spectral efficiency of each UT to all cells is fixed to a constant. Allowing the interference to be a function of the UT assignment, which allows for a more realistic construction of the problem, increases the complexity even further and thereby necessitates a heuristic method. For this case, we present a genetic algorithm based cell switch-off scheme which offers good results with linear complexity.