Inventions made in the last century laid the ground-work for the development of wireless communications; one of the largest sectors of the telecommunications industry. At present, there are more mobile connections than there are people on Earth, while wireless systems and devices such as smartphones have penetrated all sectors of the society at an unprecedented scale. As such, energy consumption has become a significant concern for green wireless systems operation [1]. Traditionally, wireless system's design has focused on performance optimization such as maximizing the spectral efficiency, throughput and minimizing the end-to-end communication latency. On the other hand, energy efficiency (EE) of wireless communications, which was mostly overlooked in the operation of previous generations of wireless systems, is now a key figure of merit [2]. Over the past few years, telecommunication operators across the world have seen their revenues eroding, while infrastructure, operation and maintenance costs have increased. At the same time, research projects have found that the information and communications technology (ICT) industry is responsible for a major percentage of greenhouse gas emissions such as carbon dioxide. As more wireless networks and devices get connected every day, pollution levels will further rise, and it is essential to reduce harmful emissions to acceptable levels in order to act on the threat of global warming and climate change.

Driven by the demand to accommodate today's growing mobile traffic, 5G is designed to be a key enabler and a leading infrastructure provider in the information and communication technology industry by supporting a variety of forthcoming services with diverse requirements. Considering the ever-increasing complexity of the network, and the emergence of novel use cases such as autonomous cars, industrial automation, virtual reality, e-health, and several intelligent applications, machine learning (ML) is expected to be essential to assist in making the 5G vision conceivable. This paper focuses on the potential solutions for 5G from an ML-perspective. First, we establish the fundamental concepts of supervised, unsupervised, and reinforcement learning, taking a look at what has been done so far in the adoption of ML in the context of mobile and wireless communication, organizing the literature in terms of the types of learning. We then discuss the promising approaches for how ML can contribute to supporting each target 5G network requirement, emphasizing its specific use cases and evaluating the impact and limitations they have on the operation of the network. Lastly, this paper investigates the potential features of Beyond 5G (B5G), providing future research directions for how ML can contribute to realizing B5G. This article is intended to stimulate discussion on the role that ML can play to overcome the limitations for a wide deployment of autonomous 5G/B5G mobile and wireless communications

The introduction of renewable energy sources (RESs) as power supply for communication systems and, for wireless cellular networks in particular, is becoming more and more attractive for a number of reasons. First, the need to reduce network operation costs through energy saving. Second, the interest in bringing cellular communications to areas of the world where the power grid is not developed and/or reliable, or in emergency situations, generates a great interest in off-grid base stations (BSs) that are energy self-sufficient. Finally, the introduction of RES as power supply is a promising way to start responding to the timely issue of Information and Telecommunication Technology (ICT) sustainability. In this chapter, we discuss the technological challenges associated with the introduction of RES-based power supply for wireless networks. Sources like photovoltaic (PV) panels and small wind turbines are the most suited ones for powering cellular access networks, due to their limited size and relatively ease of deployment. However, these sources are intermittent and generate variable amounts of energy not always easy to predict. Network operations require mechanisms and algorithms for deciding the optimal configuration that depends also on consumption and energy availability. Optimality of network operation is not simply performance maximization but becomes also consumption reduction, cost minimization, and emission reduction, through the optimal usage of the locally produced energy. In addition, considerations on the power supply dimensioning will also be presented in this chapter.

Internet-of-Things (IoT) is expected to connect tens of billions of devices anytime and anywhere and enable a wide range of services such as smart city, connected vehicles, and health care [1]. Recent advancements have driven the rapid growth of IoT in 5G technologies along with cloud- and edge-computing-enabled big-data analytics. However, one typical drawback of the existing IoT solution is the limited lifetime due to the massive number of IoT devices being powered by batteries with finite capacities. Therefore, keeping a large number of energy-constrained IoT devices alive poses a key design challenge. To this end, Backscatter Communication (BackCom) has emerged as a promising technology, allowing IoT devices to transmit data with low-power consumption. Moreover, its low-complexity design and small form factor make BackCom more attractive by realizing cost-effective IoT deployment. We organize the remainder of this chapter as follows. In Section 7.1, we provide fundamental knowledge for BackCom, including the basic principles, key design parameters, and standardization. Then, we summarize several BackCom networks in Section 7.2 and introduce several advanced emerging communication technologies redesigned for BackCom in Section 7.3. In Section 7.4, we explain several performance improvement methods of BackCom. Next, we focus on the applications empowered by BackCom in Section 7.5. Last, we discuss the open issues and future directions of BackCom in Section 7.6.

The aim of this chapter is to take a closer look at the fundamental limits of EE in massive MIMO, which is one of the key physical -layer technologies in 5G cellular networks and, more generally, to meet the ever-growing demand for wireless broadband connectivity.

Cloud radio access network (C-RAN) is an evolutionary radio network architecture in which a cloud-computing-based baseband (BB) signal-processing unit is shared among distributed low-cost wireless access points. This architecture offers a number of significant improvements over the traditional RANs, including better network scalability, spectral, and energy efficiency. As such C -RAN has been identified as one of the enabling technologies for the next-generation mobile networks. This chapter focuses on examining the energy-efficient transmission strategies of the C-RAN for cellular systems. In particular, we present optimization algorithms for the problem of transmit beamforming designs maximizing the network energy efficiency. In general, the energy efficiency maximization in C-RANs inherits the difficulty of resource allocation optimizations in interference-limited networks, i.e., it is an intractable non convex optimization problem. We first introduce a globally optimal method based on monotonic optimization (MO) to illustrate the optimal energy efficiency performance of the considered system. While the global optimization method requires extremely high computational effort and, thus, is not suitable for practical implementation, efficient optimization techniques achieving near -optimal performance are desirable in practice. To fulfill this gap, we present three low -complexity approaches based on the state-of-the-art local optimization framework, namely, successive convex approximation (SCA).

This chapter introduces the basic concepts on energy efficiency and non -orthogonal multiple access (NOMA) to unlock the potentials of future communication networks. The energy -efficient resource allocation design for NOMA systems is formulated as a non -convex optimization problem. Based on the fractional programming and successive convex approximation (SCA), a generic algorithm is proposed to achieve a suboptimal solution of the formulated problem. Simulation results are provided to verify the convergence of the proposed algorithm and to evaluate the system energy efficiency of the proposed design.

This book has provided a thorough discussion of research trends in the field of energy-efficient wireless systems and networks. Our treatment of this topic has been presented in three main parts. The first part described new mathematical tools and concepts that can be used to analyze communication systems in order that they can be configured to operate in a very energy-efficient manner. The second part of this book moved on to discuss how wireless devices and networks can be powered using either renewable energy sources, such as wind or solar power, or even by energy harvesting, where a radio receiver can exploit the energy available from the radio spectrum directly. The final part of this book explores a number of promising new technologies for energy-efficient operation, including concepts such as massive multiple input multiple output (MIMO), interference cancellation approaches, and innovative visible light communications. Ever since the 3GPP standards body defined the first new radio release 15 in mid-2018, the evolution of fifth-generation (5G) wireless technologies has progressed with new opportunities to improve the energy efficiency [1]. In this concluding chapter, we discuss some of them and take a look into several promising green solutions for the realization of 5G phase II and beyond. These points are organized into two main sections, dealing with “Flattening the Energy Curve to Support 5G Evolution” and “Potential Solutions for a Green Future".