The wireless edge caching is considered as a promising technique to cope with rapid increase in mobile traffic demand. The fundamental idea of edge caching is to offload the data traffic to local cache memories by dealing with content requests with the pre-fetched contents on network edge nodes. The wireless edge caching consists of two main phases: content placement and content delivery. Since the strategies for these two phases are highly dependent on which devices are capable of caching in the network, the characteristics and types of achievable caching gains appear to vary with the location of cached data. The cached data at the transmitter side can be utilized to reduce the traffic load on backhaul and the latency, while the cached data at the receiver side can be utilized to improve the network resource efficiency and the quality of experience (QoE) of the end-users. This chapter introduces the state-of-the-art wireless edge caching techniques for transmitters and receivers of ultra dense networks and offers a design guideline on reaping the promising gain of wireless edge caching.

Ultra dense networks with directional antennas, like millimetre wave (mmWave) networks, have some promising features about secure communications. This chapter explores the potential of physical layer security in mmWave ultra dense networks. Specifically, we mainly introduced the impact of mmWave channel characteristics, random blockages, and antenna gains on the secrecy performance. Our results reveal that mmWave frequency to high mmWave frequency is demanded to obtain a higher secrecy rate. In addition, new antenna pattern models are needed to well characterize the effective antenna gain for a random interferer seen by the typical receiver when the number of mmWave antennas grows large.

To take full advantage of the ultra-dense architecture and efficiently serve the traffic with spatiotemporal fluctuation, the transmission mechanisms should be redesigned under the constraints of backhaul and energy consumption. In this chapter, we summarize and classify the spatiotemporal arrival properties of different traffic in ultra-dense networks, and optimize several promising technologies to match the traffic. A new approach based on combining stochastic geometry and queueing theory is proposed to provide a useful guidance for the design of ultra-dense networks.

Full duplex ultra-dense network (FDUDN) is envisioned as a promising network paradigm for spectrum efficiency enhancement. This chapter presents a power management scheme, which maximizes the total capacity of FDUDN, under given Quality-of-Service (QoS) and cross-tier interference constraints. The inter-cell power control is formulated as a non-convex optimization problem and the variable substitution is used to transform it into a convex one. Furthermore, the problem is solved through a low-complexity heuristic scheme, which utilizes the water-filling theorem in inter-cell power allocation. Simulation demonstrates the enhancement effect of the proposed scheme in terms of the capacity and the power efficiency.

This chapter investigates the application non-orthogonal multiple access (NOMA) in heterogeneous ultra-dense networks (HUDNs).Particularly, we propose a unified NOMA framework first. Then the applications of the proposed unified NOMA framework in HUDNs will be discussed. With the fact that small cells are densely deployed and the non-orthogonality of resource sharing, the system suffers severe interference. In this chapter, we identify the key challenges in the unified NOMA enabled HUDNs, especially for user association and resource allocation. In addition, we carry out the related case studies for the proposed unified NOMA enabled HUDNs including the user association based on matching theory and resource allocation based on optimization techniques. Furthermore, some critical insights will be provided for the design of NOMA enabled HUDNs, which can promote network access capacity in the next generation of communication systems.

The network densification is one of the prominent solutions for fifth-generation (5G) networks to utilize spectrum resources through intensive deployment of small cells. However, the traffic management in dense networks become a serious challenge for underlying infrastructure supporting the virtual core network. Moreover, 5G will employ different types of communication frameworks: ultra-reliable low latency communication (URLLC), enhanced Mobile Broadband (eMBB), and massive Internet of Things (mIoT). Each identify standard slice type (STT) that have different performance requirements and enabling technologies. The current network developers do not provide any concise identification on how those logic networks would be administrated on top of physical network. This chapter investigates the 5G sliced networks and study virtual networking options to meet the performance requirements of service-based architecture.

Recently, with the development and popularization of unmanned aerial vehicles (UAVs), researches on UAVs have also been attracted increasing attention in wireless communications. By fully exploiting their potentials, leveraging UAVs to assist UDNs can greatly improve the system performance. The probability of having line-of-sight transmissions is higher than the probability in the terrestrial transmissions. Moreover, their attitudes can be freely adjusted. This chapter presents a vision of UAV based UDNs to exploit the potential merits of UAVs in UDNs. The channel characteristics are first discussed. Followed that, the representative scenarios where aerial UAVs are introduced to enhance terrestrial UDNs are investigated in detail, i.e., UAV-enabled BSs, UAV-enabled relays, and UAV-enabled energy transfer. UAV based UDNs also face many challenges, such as the limited spectrum resource and on-board energy. From the spectrum sharing perspective, this chapter also discusses the robust spectrum sharing optimization between UAV communications and terrestrial communications, where the scenarios with abnormal behaviors and uncertain/incomplete information are considered.

Due to the proliferation of smart devices, internet-of-things (IoT) devices, and device-to-device (D2D) communications, the amount of mobile traffic is ever-growing. To satisfy this skyrocketing wireless data traffic demand of mobile users, the densification of the network is unaviodable. However, the denser network not only improves the transmission rate, but also increases the impact of interferences. Therefore, careful deployment of base stations (BSs) is needed to guarantee the communication quality. This chapter provides insights into the deployment of BSs in the multi-layer ultra dense network (UDN). Throughout this chapter, we will model the channel between the BSs and a typical user equipment (UE) by considering the antenna height of the BSs and derive the expressions for the interference, the coverage probability, and the area spectral efficiency (ASE) of our considered multi-layer UDN using stochastic geometry. Through numerical results, we will show how the network performance can be maximized by selecting the proper antenna heights and the densities of BSs in the multi-layer UDN.

As mobile data traffic keeps growing and mobile applications pose increasingly stringent and diverse requirements, wireless networks are facing unprecedented pressures. Network infrastructure densification presents promises to further evolve wireless networks and maintain their competitiveness. Deploying more radio access points equipped with storage and computation capabilities can increase network capacity, improve network energy efficiency, provide low-latency services and access for massive devices. The benefits of network densification can be exploited using the emerging fog radio access network (Fog-RAN) architecture by pushing computation and storage resources to network edges. However, it comes with formidable technical challenges. Innovative methodologies are needed to operate such networks with various resources. This chapter develops a generalized low-rank optimization model for performance enhancements in ultra-dense Fog-RANs, supported by various motivating design objectives including mobile edge caching and topological interference alignment. A special attention is paid on algorithmic approaches for nonconvex low-rank optimization problems via Riemannian optimization.

Recently, software-defined networking (SDN) has been expected as an efficient technology to realize flexible resource management and system performance control by separating resource management from geo-distributed resources, especially for heterogeneous ultra-dense networks (HetUDNs). This work establishes an SDN based architecture for mobile traffic offloading in HetUDNs, which consist of densely deployed macro-cell base stations (MBSs) and small-cell base stations (SBSs). Additional, we explore a scenario with information asymmetric, specifically, the capacity of the SBSs can be accessible, but their performance for offloading cannot be obtained by the controller of SDN. To address such asymmetry, we propose a bundle of traffic offloading contracts, which are capable of encouraging each SBS to select the right contract that designed personally to it by promising its maximum utility. Moreover, by designing the contracts which offer rationality and incentive compatibility to different SBS types, the characteristics of a large number of SBSs are aggregated to support the efficient selection on SBSs to provide traffic offloading. Then a closed-form expression for SBS types is proposed, and we prove the monotonicity and incentive compatibility of the resulting contracts. Furthermore, simulation results validate the system performance, and the effectiveness and efficiency of the proposed contract-based traffic offloading mechanism.

This chapter studies the potential of physical layer security in future ultra-dense networks. The unique features of ultra-dense networks are summarized, which can be exploited to enhance the secure transmission at the physical layer. We illustrate that physical layer security can be implemented in many use cases such as vehicle-to-everything, edge computing and caching services, to safeguard the confidential messages. The opportunities and challenges in these research areas are presented.

In recent years, the overall network data traffic is dramatic increasing,a promising solution is the deployment of ultra dense networks (UDNs) combined with millimeter wave (mmWave) communication technology, which is expected to enhance the overall performance of the network in terms of energy efficiency and load balancing. In this chapter,user association and power allocation inmmWave-based UDNs is considered with attention to load balance constraints, energy harvesting by base stations, user quality of service requirements, energy efficiency, and cross-tier interference limits. This chapter not only establish the system utility optimal function model in the limitations of power and QoS, but also gives an iterative gradient user association and power allocation algorithm to resolve the optimization issue. This algorithm provides a best ratio of convergence and can get a near optimal scheme. In addtion, through utilizing Lagrangian dual decomposition, the dual optimization issue is disintegrated to two sub-problems, we can resolve them respectively. The simulation datum indicate that our method is effective.

Network densification has become a major contributor to expanding network capacity for the 5th generation and beyond wireless networks. Despite the potential benefits, however, the network over-densification would as well result in unhandlable interference, which is primarily due to the over-use of spectral resources. Therefore, whether the available interference management (IM) techniques are still capable of effectively handling the interference in ultra-dense networks (UDN) becomes doubtful. In this chapter, we first study the new features of the interference in UDN. Then, we make a brief overview of the IM techniques. Performance evaluation is further made, which indicates typical IM techniques fail to effectively mitigate interference in UDN. Considering the new features of the interference, we then discuss how to implement effectively interference management through designing an IM entity for UDN. With the aid of the IM entity, we tailor an effectively IM approach, which is capable of mitigating the severe interference and decorrelating the temporal interference correlation. Results show that the proposed could greatly enhancing network capacity in UDN.