Due to the extensive growth in the number of subscribers and service usage times, a notable increase in energy consumption causes more and more harmful impact on the environment and makes Information and Communication Technologies (ICTs) a major contributor to overall green house gas emissions. For instance, mobile networks represent already around 10% of the total carbon emitted by ICT and this is expected to increase every year. Several efforts were proposed to save energy in Radio Access Network (RAN) of the recent 4G Long Term Evolution (LTE) by completely switching off Base Stations (BS) during off-peak hours when data traffic is low because an active BS in an idle status (no transmission) consumes more than 50% of the energy due to circuit processing, air conditioning and other factors. On the other hand, the smart grid technology can significantly help to enhance the energy efficiency in cellular networks and contribute to the reduction of greenhouse gas emissions mainly when renewable energies are available as alternative power sources. Indeed, the smart grid is considered as a new global commercial venture widely seen as the mean to upgrade the electrical infrastructure to enhance energy savings and help to optimize some green goals of consumers by reducing house gas emissions and adjust optimally the consumed power. 

Several methods that improve the energy efficiency in LTE cellular networks by not only applying the BS sleeping strategy but also by introducing the smart grid as a tool or power management are proposed. We formulate a multi-objective optimization problem that leads to the maximization of the profit of a Long-Term Evolution (LTE) cellular operator, and at the same time to the minimization of CO2 emissions in green wireless cellular networks without affecting the desired Quality of Service (QoS). We propose two heuristic algorithms to solve the optimization problem. The first one is based on eliminating BS successively until reaching a suboptimal solution while the second is based on the Genetic Algorithm (GA). The proposed algorithms take into account both radio UpLink (UL) and DownLink (DL) directions, LTE resource allocation and intercell interference. 

In our numerical results, we analyze the performance of the proposed system versus the attitude of the mobile operator towards the environment. In fact, we introduce a Pareto weight that reflects the attitude of the operator. When this weight goes to 0, we deal with a selfish network operator that aims to maximize its own profit regardless of its impact on the environment while, when it goes to 1, we are dealing with an environmentally friendly network operator that aims to reduce CO2 emissions regardless of its own profit. Other values of the Pareto weight constitute a tradeoff between these two extremes. Furthermore, we compare the results of both proposed algorithms and we investigate two scenarios depending on how renewable energies are procured: from public retailers or from private renewable energy equipment deployed on BS sites by the mobile operator. Finally, we focus on the impact of the variation of the unitary price of renewable energies on the behavior of the mobile network towards the environment.