Due to the increase in data traffic and the number of mobile network users in this decade, the energy consumption of the radio access part of mobile networks is expected to continue to grow significantly in the upcoming years. High energy consumption of mobile networks contributes to the increase in operating costs of mobile operators as well as unwanted greenhouse gas emissions. This paper describes methods for reducing the energy consumption of the fifth (5G) and sixth (6G) mobile network generations. Special emphasis is given to the methods based on artificial intelligence, machine and deep learning. The thesis is structurally divided into six chapters. The first chapter describes the concept of artificial intelligence, as well as different types of applications of artificial intelligence such as machine and deep learning. Mobile networks with special emphasis on the 5G and 6G networks are then described in the next chapter. The energy consumption of the 5G and 6G mobile networks as well as the main causes of high energy consumption in radio access networks are analyzed. Various approaches are presented, with the aim to reduce the energy consumption of the radio part of the 5G network while maintaining the existing quality of the network service. In addition, the approaches of the largest base station manufacturers in terms of technical solutions that contribute to the reduction of network energy consumption which are implemented in existing generations of base stations are described. Furthermore, in the penultimate chapter methods for reducing the energy consumption of the 5G and 6G mobile networks based on the application of artificial intelligence, deep and machine learning were analyzed. In particular, methods based on the optimal setup of the base station activity and transmission power with the management of the mobile network resources including resource allocation in the massive multiple-input multiple-outputs (mMIMO) systems and energy harvesting from renewable energy sources are analyzed. In the last chapter is concluded that the implementation of artificial intelligence and corresponding method can be an important tool that will be predominantly used in the future to optimize energy consumption in the radio part of wireless cellular networks.