There is no doubt that big data are now rapidly expanding in all science and engineering domains. While the potential of these massive data is undoubtedly significant, fully making sense of them requires new ways of thinking and novel learning techniques to address the various challenges. In this paper, we present a literature survey of the latest advances in researches on machine learning for big data processing. First, we review the machine learning techniques and highlight some promising learning methods in recent studies, such as representation learning, deep learning, distributed and parallel learning, transfer learning, active learning, and kernel-based learning. Next, we focus on the analysis and discussions about the challenges and possible solutions of machine learning for big data. Following that, we investigate the close connections of machine learning with signal processing techniques for big data processing. Finally, we outline several open issues and research trends.

Spectrum inference, also known as spectrum prediction in the literature, is a promising technique of inferring the occupied/free state of radio spectrum from already known/measured spectrum occupancy statistics by effectively exploiting the inherent correlations among them. In the past few years, spectrum inference has gained increasing attention owing to its wide applications in cognitive radio networks (CRNs), ranging from adaptive spectrum sensing, and predictive spectrum mobility, to dynamic spectrum access and smart topology control, to name just a few. In this paper, we provide a comprehensive survey and tutorial on the recent advances in spectrum inference. Specifically, we first present the preliminaries of spectrum inference, including the sources of spectrum occupancy statistics, the models of spectrum usage, and characterize the predictability of spectrum state evolution. By introducing the taxonomy of spectrum inference from a time-frequency-space perspective, we offer an in-depth tutorial on the existing algorithms. Furthermore, we provide a comparative analysis of various spectrum inference algorithms and discuss the metrics of evaluating the efficiency of spectrum inference. We also portray the various potential applications of spectrum inference in CRNs and beyond, with an outlook to the fifth-generation mobile communications and next generation high frequency communications systems. Last but not least, we highlight the critical research challenges and open issues ahead.

In this paper, we consider the spectrum sharing planning problem for full-duplex unmanned aerial vehicle (UAV) relaying systems with underlaid device-to-device (D2D) communications, where a mobile UAV employed as a full-duplex relay assists the communication link between separated nodes without direct link. Our design aims to maximize the sum throughput under the transmit power budget, while guaranteeing the coexistence with terrestrial D2D pairs, satisfying the information causality and UAV’s trajectory constraints. Firstly, the transmit power planning with a given trajectory is investigated, where a successive convex algorithm is developed by leveraging the D.C. (difference of two convex) programming. Then, we propose a two-step trajectory design method for the given transmit power since the constraints of D2D pairs result in a non-convex feasible set. Further, an efficient spectrum sharing method for aerial UAV and terrestrial D2D communications is designed by alternately optimizing the transmit power and UAV’s trajectory. Finally, simulation results under various parameter configurations are provided to show the effectiveness of the proposed algorithms.

In this paper, we investigate the resource allocation problem for unmanned aerial vehicle (UAV)-assisted networks, where a UAV acting as an energy source provides radio frequency energy for multiple energy harvesting-powered device-to-device (D2D) pairs with much information to be transmitted. The goal is to maximize the average throughput within a time horizon while satisfying the energy causality constraint under a generalized harvest-transmit-store model, which results in a non-convex problem. By introducing the Lagrangian relaxation method, we analytically show that the behavior of all D2D pairs at each time slot is exclusive: harvesting energy or transmitting information signals. The formulated non-convex optimization problem is thus transformed into a mixed integer nonlinear programming (MINIP). We then design an efficient resource allocation algorithm to solve this MINIP, where D.C. (difference of two convex functions) programming and golden section method are combined to achieve a suboptimal solution. Furthermore, we provide an idea to reduce the computational complexity for facilitating the application in practice. Simulations are conducted to validate the effectiveness of the proposed algorithm and evaluate the system throughput performance.

This letter investigates the transmit power and trajectory optimization problem for unmanned aerial vehicle (UAV)-aided networks. Different from majority of the existing studies with fixed communication infrastructure, a dynamic scenario is considered where a flying UAV provides wireless services for multiple ground nodes simultaneously. To fully exploit the controllable channel variations provided by the UAV’s mobility, the UAV’s transmit power and trajectory are jointly optimized to maximize the minimum average throughput within a given time length. For the formulated non-convex optimization with power budget and trajectory constraints, this letter presents an efficient joint transmit power and trajectory optimization algorithm. Simulation results validate the effectiveness of the proposed algorithm and reveal that the optimized transmit power shows a water-filling characteristic in spatial domain.