Joint management of energy consumption, maintenance costs and user revenues in cellular networks with sleep modes

We target the maximization of the operator profitability in a LTE cellular network with e-NodeBs (eNBs) exploiting Sleep Mode (SM) power states. The profitability is composed of different terms, namely: the electricity bill due to the eNBs energy consumption, the maintenance costs due to the application of power states to the eNBs, and the revenues from providing a given throughput to users. After showing that all these components are strictly interdependent, we formulate the problem of maximizing the profitability for a set of eNBs. In addition, we propose an algorithm, called EMRA, as a practical solution of the optimization problem. Results, obtained over different cellular scenarios, prove that EMRA outperforms previous algorithms, which instead focus solely on the energy bill, the maintenance costs or the revenues in isolation, i.e., without considering them all together. Moreover, EMRA can be adapted to different network conditions and to different scenarios, bringing always a revenue for the network operator.