The rapid uptake of mobile devices and the rising popularity of mobile applications and services pose unprecedented demands on mobile and wireless networking infrastructure. Upcoming 5G systems are evolving to support exploding mobile traffic volumes, real-time extraction of fine-grained analytics, and agile management of network resources, so as to maximize user experience. Fulfilling these tasks is challenging, as mobile environments are increasingly complex, heterogeneous, and evolving. One potential solution is to resort to advanced machine learning techniques, in order to help manage the rise in data volumes and algorithm-driven applications. The recent success of deep learning underpins new and powerful tools that tackle problems in this space. In this paper, we bridge the gap between deep learning and mobile and wireless networking research, by presenting a comprehensive survey of the crossovers between the two areas. We first briefly introduce essential background and state-of-the-art in deep learning techniques with potential applications to networking. We then discuss several techniques and platforms that facilitate the efficient deployment of deep learning onto mobile systems. Subsequently, we provide an encyclopedic review of mobile and wireless networking research based on deep learning, which we categorize by different domains. Drawing from our experience, we discuss how to tailor deep learning to mobile environments. We complete this survey by pinpointing current challenges and open future directions for research.

/pdf/deep-learning-in-mobile-and-wireless-networking-a-survey-ct8cy2kuqi.pdf

Deep Learning in Mobile and Wireless Networking: A Survey

The exponential growth and availability of data in all forms is the main booster to the continuing evolution in the communications industry. The popularization of traffic-intensive applications including high definition video, 3-D visualization, augmented reality, wearable devices, and cloud computing defines a new era of mobile communications. The immense amount of traffic generated by today’s customers requires a paradigm shift in all aspects of mobile networks. Ultradense network (UDN) is one of the leading ideas in this racetrack. In UDNs, the access nodes and/or the number of communication links per unit area are densified. In this paper, we provide a survey-style introduction to dense small cell networks. Moreover, we summarize and compare some of the recent achievements and research findings. We discuss the modeling techniques and the performance metrics widely used to model problems in UDN. Also, we present the enabling technologies for network densification in order to understand the state-of-the-art. We consider many research directions in this survey, namely, user association, interference management, energy efficiency, spectrum sharing, resource management, scheduling, backhauling, propagation modeling, and the economics of UDN deployment. Finally, we discuss the challenges and open problems to the researchers in the field or newcomers who aim to conduct research in this interesting and active area of research.

Ultra-Dense Networks: A Survey

The fifth generation (5G) mobile networks are envisioned to support the deluge of data traffic with reduced energy consumption and improved quality of service (QoS) provision. To this end, key enabling technologies, such as heterogeneous networks (HetNets), massive multiple-input multiple-output (MIMO), and millimeter wave (mmWave) techniques, have been identified to bring 5G to fruition. Regardless of the technology adopted, a user association mechanism is needed to determine whether a user is associated with a particular base station (BS) before data transmission commences. User association plays a pivotal role in enhancing the load balancing, the spectrum efficiency, and the energy efficiency of networks. The emerging 5G networks introduce numerous challenges and opportunities for the design of sophisticated user association mechanisms. Hence, substantial research efforts are dedicated to the issues of user association in HetNets, massive MIMO networks, mmWave networks, and energy harvesting networks. We introduce a taxonomy as a framework for systematically studying the existing user association algorithms. Based on the proposed taxonomy, we then proceed to present an extensive overview of the state-of-the-art in user association algorithms conceived for HetNets, massive MIMO, mmWave, and energy harvesting networks. Finally, we summarize the challenges as well as opportunities of user association in 5G and provide design guidelines and potential solutions for sophisticated user association mechanisms.

/pdf/user-association-in-5g-networks-a-survey-and-an-outlook-4zdf9nh1ox.pdf

User Association in 5G Networks: A Survey and an Outlook

Fueled by the availability of more data and computing power, recent breakthroughs in cloud-based machine learning (ML) have transformed every aspect of our lives from face recognition and medical diagnosis to natural language processing. However, classical ML exerts severe demands in terms of energy, memory, and computing resources, limiting their adoption for resource-constrained edge devices. The new breed of intelligent devices and high-stake applications (drones, augmented/virtual reality, autonomous systems, and so on) requires a novel paradigm change calling for distributed, low-latency and reliable ML at the wireless network edge (referred to as edge ML). In edge ML, training data are unevenly distributed over a large number of edge nodes, which have access to a tiny fraction of the data. Moreover, training and inference are carried out collectively over wireless links, where edge devices communicate and exchange their learned models (not their private data). In a first of its kind, this article explores the key building blocks of edge ML, different neural network architectural splits and their inherent tradeoffs, as well as theoretical and technical enablers stemming from a wide range of mathematical disciplines. Finally, several case studies pertaining to various high-stake applications are presented to demonstrate the effectiveness of edge ML in unlocking the full potential of 5G and beyond.

/pdf/wireless-network-intelligence-at-the-edge-36kakqk6fq.pdf

Wireless Network Intelligence at the Edge

Given the unprecedented availability of data and computing resources, there is widespread renewed interest in applying data-driven machine learning methods to problems for which the development of conventional engineering solutions is challenged by modeling or algorithmic deficiencies. This tutorial-style paper starts by addressing the questions of why and when such techniques can be useful. It then provides a high-level introduction to the basics of supervised and unsupervised learning. For both supervised and unsupervised learning, exemplifying applications to communication networks are discussed by distinguishing tasks carried out at the edge and at the cloud segments of the network at different layers of the protocol stack, with an emphasis on the physical layer.

/pdf/a-very-brief-introduction-to-machine-learning-with-507l7aokj4.pdf

A Very Brief Introduction to Machine Learning With Applications to Communication Systems

In this article, we use deep neural networks (DNNs) to develop an end-to-end wireless communication system, in which DNNs are employed for all signal-related functionalities, including encoding, decoding, modulation, and equalization. However, accurate instantaneous channel transfer function, i.e., the channel state information (CSI), is necessary to compute the gradient of the DNN representing. In many communication systems, the channel transfer function is hard to obtain in advance and varies with time and location. In this article, this constraint is released by developing a channel agnostic end-to-end system that does not rely on any prior information about the channel. We use a conditional generative adversarial net (GAN) to represent the channel effects, where the encoded signal of the transmitter will serve as the conditioning information. In addition, in order to obtain accurate channel state information for signal detection at the receiver, the received signal corresponding to the pilot data is added as a part of the conditioning information. From the simulation results, the proposed method is effective on additive white Gaussian noise (AWGN) and Rayleigh fading channels, which opens a new door for building data-driven communication systems.

Channel Agnostic End-to-End Learning Based Communication Systems with Conditional GAN

Channel state information (CSI) at the base station (BS) is critical to resource allocation design for wireless networks, but it is hard to obtain accurate CSI in a high mobility vehicular environment. In this letter, we study the spectrum and power allocation problem in device-to-device-enabled vehicular networks, where CSI of vehicular links is only reported to the BS periodically. We maximize the sum throughput of all vehicle-to-infrastructure (V2I) links while guaranteeing the reliability of each vehicle-to-vehicle (V2V) link with the delayed CSI feedback. We propose a low-complexity algorithm to find the optimal spectrum sharing strategy among V2I and V2V links and properly adjust their transmit powers. Its desirable performance is confirmed by computer simulation.

Spectrum and Power Allocation for Vehicular Communications With Delayed CSI Feedback

In this article, we use deep neural networks (DNNs) to develop a wireless end-to-end communication system, in which DNNs are employed for all signal-related functionalities, such as encoding, decoding, modulation, and equalization. However, accurate instantaneous channel transfer function, \emph{i.e.}, the channel state information (CSI), is necessary to compute the gradient of the DNN representing. In many communication systems, the channel transfer function is hard to obtain in advance and varies with time and location. In this article, this constraint is released by developing a channel agnostic end-to-end system that does not rely on any prior information about the channel. We use a conditional generative adversarial net (GAN) to represent the channel effects, where the encoded signal of the transmitter will serve as the conditioning information. In addition, in order to deal with the time-varying channel, the received signal corresponding to the pilot data can also be added as a part of the conditioning information. From the simulation results, the proposed method is effective on additive white Gaussian noise (AWGN) and Rayleigh fading channels, which opens a new door for building data-driven communication systems.

Channel Agnostic End-to-End Learning based Communication Systems with Conditional GAN.

Dual Connectivity in LTE network can significantly improve per-user throughput and mobility robustness by allowing users to be connected simultaneously to master cell group (MCG) and secondary cell group (SCG) via MeNB (master eNB) and SeNB (secondary eNB), respectively. The increase in per-user throughput is achieved by aggregating radio resources from at least two eNBs. Moreover, dual connectivity also helps in load balancing between MCG and SCG. However, it imposes several technical challenges. The main ones are buffer status report calculation and reporting, power headroom calculation and reporting, logical channel prioritization, user power saving operations such as discontinuous reception (DRX), and increased device complexity to support bearer split. The coordination between eNBs over X2 interface may be effective to resolve some of these issues. The higher delay due to non-ideal backhaul between MeNB and SeNB, however, limits an efficient coordination between these eNBs. In this paper, we explain and explore these technical challenges and investigate potential solution directions. We also provide a quantitative analysis of potential gains in terms of peruser throughput and load balancing that can be achieved by data bearer split in uplink at the cost of complex UE behaviors using a system level simulation study.

Dual Connectivity in LTE small cell networks

Technology for communicating data to a user equipment (UE) is disclosed. Effective data rates may be identified, at a master evolved node B (MeNB), for the UE with respect to the MeNB and a secondary evolved node B (SeNB). A downlink split ratio for bearers of the MeNB and the SeNB to the UE may be determined based in part on the effective data rates for the UE. A first portion of data may be sent from the MeNB to the UE according to the downlink split ratio. A remaining portion of data may be sent from the MeNB to the UE via the SeNB according to the downlink split ratio, wherein the UE supports dual connectivity with the MeNB and the SeNB.

Kathiravetpillai Sivanesan

Papers

Channel Agnostic End-to-End Learning Based Communication Systems with Conditional GAN

Spectrum and Power Allocation for Vehicular Communications With Delayed CSI Feedback

Channel Agnostic End-to-End Learning based Communication Systems with Conditional GAN.

Dual Connectivity in LTE small cell networks

Bearer split ratios for dual connectivity systems