In writing this text, we hope to achieve multiple goals. Firstly, we hope to develop a textbook that is useful as a reference for graduate or a supplement to advanced undergraduate classes investigating advanced wireless communications. These topics include adaptive antenna processing, multiple-input multiple-output (MIMO) communications, and wireless networks. Throughout the text, there is a recurring theme of understanding and mitigating both internal and external interference. In addressing these areas of investigation, we explore concepts in information theory, estimation theory, signal processing, and implementation issues as are applicable. We attempt to provide a development covering these topics in a reasonably organized fashion. While not always possible, we attempt to be consistent in notation across the text. In addition, we provide problem sets that allow students to investigate these topics more deeply. Secondly, we attempt to organize the topics addressed so that this text will be useful as a reference. To the extent possible, each chapter will be reasonably self-contained, although some familiarity with the topic area is assumed. To aid the reader, reviews of many of the mathematical tools needed within the text are collected in Chapters 2 and 3. In addition, an overview of the basics of communications theory is provided in Chapters 4 and 5. Finally, in discussing these topics, we attempt to address a wide range of perspectives appropriate for the serious student of the area. Topics range from information theoretic bounds, to signal processing approaches, to practical implementation constraints.

Point-to-point links and networks

The simplest wireless communication link is between a single transmitter and a single receiver. In point-to-point systems, data communication rates depend on factors such as bandwidth, signal power, noise power, acceptable bit-error rate, and spatial degrees of freedom.

Many wireless systems, however, comprise multiple interacting links. The parameters and trade-offs associated with point-to-point links hold for networks as well. Additional factors play a role in networks, however. For instance, interference between links can reduce data communication rates. An exciting possibility is for nodes to cooperate and help convey data for each other, which has the potential to increase data communication rates. Table 13.1 summarizes some of the key common and differentiating features of point-to-point links versus networks.

In this chapter, we analyze the performance of various multiantenna approaches in the context of cellular networks whereby signal and interference strengths are influenced by the spatial distribution of nodes and base stations. Note that we use the term cellular in a broader context than many works in the literature, which refer specifically to mobile telephone systems. Here we consider any kind of network with one-to-many (downlink) and many-to-one topologies (uplink). For most of this chapter except for Section 13.5.1, we shall focus on characterizing systems without out-of-cell interference, whereby we assume that there is some channel allocation mechanism with a reuse factor that results in negligible out-of-cell interference.

For better or worse, wireless communications have become integrated into many aspects of our daily lives. When communication systems work well, they almost magically enable us to access information from distant, even remote, sources. If one were to take a modern “smart” phone a couple of hundred years into the past, one would notice a couple of things very quickly. First, most of the capability of the phone would be lost because a significant portion of the phone's capabilities are based upon access to a communications network. Second, being burned at the stake as a witch can make for a very bad day.

There are many texts that present the history of wireless communications in great detail, for example in References [186, 48, 146, 304, 61]. Many of the papers of historical interest are reprinted in Reference [348]. Because of the rich history of wireless communications, a comprehensive discussion would require multiple texts on each area. Here we will present an abridged introduction to the history of wireless communications, focusing on those topics more closely aligned with the technical topics addressed later in the text, and we will admittedly miss numerous important contributors and events.

The early history of wireless communications covers development in basic physics, device physics and component engineering, information theory, and system development. Each of these aspects is important, and modern communication systems depend upon all of them. Modern research continues to develop and refine components and information theory. Economics and politics are an important part of the history of communications, but they are largely ignored here.

Introduction

The rapid rate of increase in data traffic means that future wireless networks will have to support a large number of users with high data rates. A promising way to achieve this is by spectrum reuse through the deployment of cells with small range, such that the same time-frequency resources may be reused simultaneously in multiple cells. At the same time, the traditional coverage requirement for wireless users (supporting a modest rate at cell-edge users) is most economically met with cells having large range, i.e., the traditional macrocellular architecture. Thus the wireless cellular networks of the future are likely to be heterogeneous, i.e., have one or more tiers of small cells overlaid on the macrocellular tier.

We consider the problem of network design from the point of view of a service provider considering the deployment of a network in a certain region. The primary metric we shall focus on is the coverage on the downlink: namely, the probability that a user at an arbitrary location in the deployed network has coverage. Later, we shall briefly discuss the distribution of the maximum rate that such a user could support from one of the base stations (BSs).

When we talk of the service provider designing a network to satisfy a certain coverage criterion, we mean the choice of deployment parameters for such a network, including:

1. Number of tiers of the network

2. Densities of the BSs in the tiers

3. Transmit powers of the BSs in the tiers.

Introduction

We are currently witnessing an exponentially increasing demand for wireless data services, which is mainly driven by the growing popularity of wireless modems, smartphones, and tablet personal computers (PCs). Not surprisingly, current cellular networks have already started reaching their capacity limits in densely populated areas and it is therefore necessary to assess which network architecture is most suited to carry the future data traffic. As additional spectrum resources are scarce, it seems inevitable that any future system architecture will rely to a significant extent on network densification, i.e., an increase of the number of antennas deployed per unit area. This can be achieved by either increasing the number of antennas per base station (BS) [1] or by deploying more BSs [2], or a combination of both [3]. More antennas per device lead to additional degrees of freedom, which can provide spatial multiplexing and diversity gains or can be used to cancel interference [4]. On the other hand, a denser deployment of BSs, such as femto or small cells [5], increases the spatial reuse of the radio spectrum. Although the network capacity would theoretically scale linearly with the BS density, dense networks suffer from increased inter-cell interference and user mobility becomes difficult to manage. The cooperation of multiple BSs, which jointly process user data from multiple cells [6] has shown its potential to counter inter-cell interference and to improve the cell edge coverage not only in theory [7] but also in practice [8]. Base station cooperation is therefore already considered as an essential feature of future cellular standards [9]. Apart from that, clusters of cooperating BSs forming virtual cells could also potentially reduce the amount of handover signaling between cells to enable user mobility in dense networks [10].

Mobile networks have undergone a major transformation in recent years, shifting from primarily delivering voice and text services to transporting data and connecting to the Internet. In many mobile networks today, data traffic already constitutes more than 97% of the total bits transmitted. Demand for data continues to be strong, driven by high take up and ease of use of smartphone, tablets, and related devices.

Long term evolution (LTE) promises one answer to the impending data capacity crunch, with lightening fast data rates and impressive spectral efficiency. However, this new technology – even with associated new spectrum – won't come close to satisfying forecast demand.

The industry now recognizes that the greatest capacity gains will be achieved by spectrum reuse through deployment of large numbers of small cells. A variety of radio technologies including 3rd Generation (3G), (LTE), and wireless fidelity (WiFi) will be relevant depending on device capabilities, spectrum availability, and price point.

No longer is femtocell technology considered purely as a coverage solution for the home. Today, it is difficult to find a network operator that does not have small cells somewhere on their roadmap.

Over the next five years, we can expect to see a major shift in network equipment investment. Analyst research already points to the majority of new cell-site investment moving across from macrocells into small cells in that timeframe. With some $50 billion of CapEx spent annually on mobile access networks, that will create both significant industry re-alignment and tremendous opportunity.

Introduction and prior art

Frequency allocation for heterogeneous networks

In heterogeneous networks, frequency resources can be allocated to different tiers in a co-channel (shared-spectrum) or dedicated channel (split-spectrum) manner, or through a hybrid technique, which is a combination of the two approaches. In the co-channel approach shown in Figure 15.1(a), while the spectrum resources are fully reused in different tiers, cross-tier interference may cause crucial setbacks to the system. For example, macrocell users in the vicinity of closed subscriber group (CSG) femtocells are not allowed to connect to the femtocells, even if their link quality with these femtocells is good. Therefore, such macrocell users receive strong downlink interference from CSG femtocells and may fall into outage.

The split spectrum approach shown in Figure 15.1(b), on the other hand, partitions the allocated spectrum between multiple tiers. Each tier can use its own segment of resource and therefore there is no cross-tier interference [1]. However, the amount of bandwidth available to each tier is reduced. Hybrid methods as shown in Figure 15.1(c) use a mixture of co-channel and dedicated channel methods, and aim to reuse the spectrum resources whenever feasible. For example, in [2], the macrocell users are dedicated a component carrier (CC), referred as the “escape carrier,” which is not used by the femtocell network. Any macrocell mobile station (MMS) that is close by to a femtocell is scheduled within this escape carrier, if the interference observed from the femtocell network is above a threshold.

Introduction

Because of the use of common radio resources in small cell networks [1, 2], destructive interference may occur not only between the femtocell and the macrocell, but also among femtocells whose coverage areas are highly overlapped with each other (collocated femtocells), as shown in Figure 10.1. To avoid interference, one typical solution is to divide the entire available spectrum into several frequency bands, and then each femtocell and the macrocell utilize different frequency bands from each other [3]. This deployment is referred to as “dedicated channel” deployment in 3rd Generation Partnership Project (3GPP) [4]. However, the performance of this solution is limited by the assigned bandwidth, which makes it infeasible for dense femtocell deployments where each femtocell can only utilize a very limited fraction of the bandwidth. An alternative solution is to adopt spatial domain frequency reuse [5]; however, this is infeasible for user deployed femtocells without centralized and perfect planning. As a result, a practical solution turns out to be “co-channel” deployment, where all femtocells and the macrocell can utilize all the available spectrum. To mitigate interference in co-channel deployment, dynamic power adaptation in femtocells has been proposed for code division multiple access (CDMA) systems [6-11] to combat interference due to the near–far problem. Considering that orthogonal frequency division multiple access (OFDMA) has been adopted by 3GPP LTE-Advanced (LTE-A) and wireless interoperability for microwave access (WiMAX), new interference mitigation techniques are needed [12]. In OFDMA, the major cause of interference is that multiple networks occupy the same radio resources (subcarriers and orthogonal frequency division multiplexing (OFDM) symbols) simultaneously.

Game theory (GT) is a mathematical tool that analyzes interactions among decision makers. Game theory is seen as a natural paradigm to study and analyze wireless networks where players compete for the same resources. The importance of studying the coexistence between macrocells and femtocells from a game theoretical perspective is multi-fold. First, as illustrated in Figure 11.1, by modeling the dynamic spectrum sharing among network players (macrocell base stations (MBSs), femtocell base stations (FBSs), mobile user equipment (MUE), and home user equipment (HUE)) as games, the behaviors and actions of players can be analyzed in a formalized structure, by which the theoretical achievements in GT can be fully utilized. Second, GT equips us with different optimality criteria for various spectrum sharing problems, which are of key importance when it comes to analyzing the equilibrium of the game. Third, the application of GT enables us to derive efficient distributed algorithms for self-organized networks relying only on partial information. In order to achieve this, the theory of strategic reinforcement learning is of utmost importance by allowing players to choose their optimal strategies and gradually learn from their environment through trial and error procedures. A comprehensive source of game theoretic approaches and their application to wireless communications can be found in [1].

Femto-macro interference coordination problem

Femtocells are small wireless access points that are typically installed in a subscriber's premises, but operate in a cellular provider's licensed spectrum. Since femtocells can be manufactured at a very low cost, require minimal network maintenance by the operator, and can leverage the subscriber's backhaul, femtocells offer the possibility of expanding cellular capacity at a fraction of the cost of traditional macrocellular deployments. With the surge in demand for wireless data services, femtocells have thus attracted considerable recent attention, both in cellular standards bodies such as the 3rd Generation Partnership Project (3GPP) [1–3] and academic research [4, 5].

However, one of the key technical challenges in deploying femtocells is the interference between the underlay of small femtocells and the overlay of comparatively large macrocells – an issue raised in virtually every survey on femtocells [5–8]. While interference is a fundamental challenge in any cellular system, the so-called cross-tier interference in femtocell networks has two particularly challenging aspects:

1. Strong and varied interference: Due to closed access or restricted association, mobile terminals (or user equipment (UE) in 3GPP terminology) may not be able to connect to a given femtocell even when it provides the closest serving base station [9]. Such restrictions can result in strong interference both from the macrocell UE transmitter onto the femtocell uplink and from the femtocell downlink onto the macrocell UE receiver. In addition, since the femtocell access points are often deployed in an essentially ad hoc manner, interference conditions are much more varied than traditional planned macrocellular networks.

User-deployed femtocells, each exclusively serving a set of registered users and sharing the same frequency spectrum as the overlay macrocells, are already defined in 3rd Generation Partnership Project (3GPP) specifications. Such a co-channel and random deployment of femtocells can cause heavy downlink (DL) control and data channel interference especially to mobile user equipment (MUE) in the vicinity of one or more femtocells and not belonging to their closed subscriber groups (CSGs). This chapter is dedicated to addressing this issue, termed as inter-cell interference coordination (ICIC), paying particular attention to the control channel. In systems that employ full frequency reuse, such as long term evolution (LTE), the issue of inter-cell interference (ICI) is a very serious one and can severely compromise cell-edge performance. The situation is further exacerbated in systems with femtocells randomly distributed within the underlying macrocellular network. In such a system, ICI is not only experienced by MUEs at the edge of macrocells, but can also be experienced by those MUEs in the vicinity of one or more femtocells, whose CSGs they are not members of. While scheduling strategies do not come under the purview of LTE standardization, LTE does provide standardized signaling methods so that an appropriate signaling strategy may be employed to avoid excessive ICI for the data channels. However, these signaling methods are developed to be exchanged between macro base stations (BSs) over the X2 interface. It is expected that LTE femtocells will not have access to such an interface. Furthermore, unlike the data channel, the various control channels cannot be conveniently relocated in order to avoid interference.

Introduction

Femtocell access points (FAPs) are foreseen to play a key role in the development and deployment of future orthogonal frequency division multiple access (OFDMA)-based cellular networks, e.g., Long Term Evolution (LTE) [1] and Wireless Interoperability for Microwave Access (WiMAX) [2], since they may deliver improved indoor coverage and network capacity [3]. Femtocell access points are low-cost, low-power, user-deployed small base stations (BSs), which provide wireless coverage of a cellular standard, and are connected to the network operator via a broadband connection, e.g., digital subscriber line (DSL), fiber optics, etc. Femtocells, as explained in the introductory chapter, offer a large number of advantages to future cellular networks. They may enhance indoor coverage, deliver both high data rates and new applications to users, and offload traffic from existing macrocell networks [4]. However, since FAPs are expected to be deployed in large numbers and because they may be installed by users in an uncoordinated manner, including self-organizing network (SON) capabilities in FAPs may be a key aspect for their successful operation of these devices [5].

A SON, defined as a network that requires minimal human involvement due to the automatic and/or autonomous nature of its functioning, integrates the processes of planning, configuration, optimization, and healing in a set of in-built automatic/autonomous functionalities. By using SON capabilities, operator intervention for network operation and maintenance can be reduced, thus minimizing deployment and operational costs of future cellular networks, which are major concerns of current mobile operators.

Introduction

Small cell networks (SCNs) made of portable pico and femto base stations (BSs) serve dense urban areas, commercial and office spaces, hotspots, etc. Their design and deployment pose many new challenges to the optimal system design. Managing mobile users deriving service from such SCNs is one of the key challenges. Furthermore, reducing cell size increases the frequency of handovers, which results in an increased number of call drops before completing the service. However, they also offer better communication rates to cell-edge users resulting in reduced service times. The study of such tradeoffs is an important topic while designing optimal systems.

In order to prevent large numbers of handovers that would result from reducing cell size, it has been proposed (for e.g., see [1, 2]) to group together a number of small cells (SCs) into one virtual macrocell. This helps to restrict the effort of preventing losses due to the handover only to those handovers that occur between SCs of the same virtual cell. In between the SCs some fast switching mechanisms are proposed such as frequency following mechanisms where the frequency used by a mobile follows it from one SC to the next. This requires reserving the same channel for a user in the entire macrocell.

In this chapter, we consider a large macrocell divided into a number of SCs and study the impact of mobility in such systems, especially the effect of frequent handovers. We assume that an ongoing call is never dropped at the SC boundary within a virtual cell (in later sections, we relax this assumption and study handover at SC boundaries within a virtual cell).