Introduction

Heterogeneous cellular network (HCN) deployments may imply an order of magnitude more network nodes than conventional homogenous macrocell deployments. Therefore, it is important that the integration and operation of these new nodes require minimal manual efforts from operators. Self-organizing network (SON) features can be seen as essential enablers to facilitate service as well as network deployment and management. The main objectives of SONs are to reduce the deployment costs, simplify network management (managing a plethora of radio access technologies (RATs) without significantly increasing operational expenses) and enhance network performance.

Within the 3rd Generation Partnership Project (3GPP) Long Term Evolution (LTE), SON was among the early system requirements, and SON features were already included in the first 3GPPLTErelease, i.e., Release 8 [1]. SONwork items in 3GPP [2, 3] have been inspired by the SONstudies and the set of requirements defined by the operators' alliance Next Generation Mobile Networks (NGMN) [4]. This chapter addresses HCN aspects of SON, although these automation features are applicable to other types of network deployment aswell. Themain focus is on LTE, but Universal Mobile Telecommunication System (UMTS) and multiple RATs will also be considered where applicable. More general discussions about SON can be found in [5–9], and a discussion with special focus on femtocells can be found in [10].

SON operations are supported by the operation, administration, and maintenance (OAM) architecture, which is presented in Section 6.2, and are commonly divided into four key components/phases: planning, self-configuration, self-optimization and selfhealing, as illustrated in Fig. 6.1.

Gordon Mansfield

Small Cell Forum

My name is Gordon Mansfield, and I currently serve as the elected chairman of the Small Cell Forum. The Forum is an industry body that promotes and drives the wide-scale adoption of small cell technologies to improve coverage, capacity and services delivered by mobile networks. I have many years of experience in the space, having previously served on the Femto Forum board from 2008-2010 and having led a tier one operators small cell effort since 2007. I consider it a great honor to be asked to write the foreword for this very informative book on small cells and heterogeneous networks. The authors are all highly respected researchers in academia and in industry, who have spent years working on the topics covered.

In recent years, small cells have become a very big topic when discussing mobile Internet and the tremendous data growth experienced over the past five years by operators around the globe. When we look at the recent history of data growth, some operators have experienced a 20,000 percent growth in data from 2007-2011. Combine that with the incredible forecast coming from all parts of the industry suggesting 10X and higher growth over the next four to five years, and it becomes clear that new ways to serve this data growth are necessary. We cannot continue to rely on new spectrum and advances in the air interface alone to sustain these types of data growth.

Introduction

The objectives behind this chapter are twofold. One is to provide an in-depth description of access control in heterogeneous cellular networks (HCNs) from the perspectives of the core network (CN), the radio access network (RAN) and the user equipment (UE). The second objective is to provide benefits and tradeoff analysis of different access control schemes through numerical simulations with respect to various performance metrics such as the percentage of offloaded UEs as well as cell-average and cell-edge data throughput.

A rudimentary understanding of the Universal Mobile Telecommunication System (UMTS) and Long Term Evolution/System Architecture Evolution (LTE/SAE) cellular architectures is essential to get some intuition behind how access control is implemented in contemporary HCNs. Section 4.2 introduces the motivation for access control and describes available access methods. Sections 4.3 and 4.4, respectively, provide a basic overview of the UMTS and Long Term Evolution (LTE) cellular architectures. We describe two main components of the system architecture, namely the CN and the RAN. At a high level, the CN is responsible for overall control of the UE including packet processing, quality of service (QoS) enforcement and connection with the operator network.The RAN is responsible for the radio air interface functions to the UE. In UMTS, the RAN functions are split between a radio network controller (RNC) and a NodeB, in LTE, the RAN functions reside at a single logical node called the evolved NodeB (eNB).

Introduction

In order to aid vendors and operators in the development and deployment of new heterogeneous cellular networks (HCNs), and the refinement of existing procedures such as handover (HO) and radio resource management (RRM), network simulation, planning and optimization tools that are able to evaluate the overall performance of complex cellular networks are highly regarded. In this context, system-level simulations have become a widely adopted methodology. In system-level simulations, the elements and operations of a cellular network are modeled by computer software. This approach is usually simpler and cheaper than real implementation, and is more accurate and reliable than analytical modeling. The number of assumptions and simplifications made in system-level simulations is up to the software designer, but it is usually less than that of analytical modeling. System-level simulations can also model more complex cellular networks than analytical modeling. However, system-level simulations for modeling intricate procedures or getting statistically representative results of network performance usually require significant computing capabilities. As a result, in order to avoid prohibitive computational costs, a tradeoff between accuracy and complexity should be reached. In this line, the 3rd Generation Partnership Project (3GPP) provides guidelines on network simulations, and defines simulation procedures and parameters.

Introduction

Compared with current cellular networks, next generation mobile networks are expected to encompass more sophisticated features, including the support of higher data transmission rates and user equipment (UE) mobility, location management, diversified service levels, etc. In order to accommodate these requirements, the 3rd Generation Partnership Project (3GPP) is devoted to the standardization of Long Term Evolution (LTE) and LTE-Advanced systems, which have been recognized as major candidates for the fourth-generation (4G) mobile networks. In LTE/LTE-Advanced systems, the network structure will be heterogeneous. How to maintain and improve mobility, handover (HO), and location management, while avoiding user experience deterioration, is a challenging task. In this chapter, we will study the mobility management challenge and illustrate advanced mobility management schemes.

In LTE/LTE-Advanced systems, the factors that make mobility, HO, and location management a challenging task are as follows

• The rapid evolution of cellular networks results in the coexistence of multiple radio access technologies (RATs), e.g., Global System for Mobile Communications (GSM), Universal Mobile Telecommunication System (UMTS) and LTE/System Architecture Evolution (SAE). This demands optimized cooperation among multiple RATs to enable UEs to roam from one RAT to another.

• The introduction of low-power nodes (LPNs) largely increases the total number of base stations (BSs), making the network structure and interference conditions more intricate. Thus, traditional mobility load balancing (MLB) and mobility management schemes need to be revisited to suit the new heterogeneous cellular network (HCN) architecture.

• The complexity of LTE/LTE-Advanced systems leads to a large number of network parameters. Therefore, efforts need to be made in defining proper key performance indicators and developing optimization techniques for mobility management in various scenarios.

Relaying is a well known technique to transmit signals from a source to a destination through one or several intermediate nodes (i.e., relay nodes (RNs)) without using increased power at the source [1–4]. In the past decade, many research efforts on relay technologies have been made to improve the cell coverage, enhance the transmission reliability, and increase the system throughput. More recently, RNs have become an important component in a heterogeneous cellular network (HCN) to provide service improvement and coverage extension at hotspots and cell edges. Layer three (L3) RN, which works as an independent base station (BS) except for the use of the wireless backhaul link, is specified in 3rd Generation Partnership Project (3GPP) Release 10 to realize flexible network deployment and increase network throughput without any additional infrastructure.

In terms of data forwarding, four types of relay have been widely studied, which are amplify-and-forward (AF), demodulate-and-forward (DMF), decode-and-forward (DCF), and estimate-and-forward (EF). Because there is no baseband signal processing function, an AF-relay-based wireless network is cost efficient. The main disadvantage of the AF relay is that the received noise and interference would also be forwarded to the destination. To mitigate the received noise and interference at an RN, the DMF relay and EF relay have been proposed to perform some simple signal processing according to the constellation used. With the decoding operation performed at the RN, a DCF relay can regenerate the source-transmitted signal perfectly if the received signal is decoded correctly.

Introduction

Nowadays, 50% of phone calls and 70% of data services are carried out indoors [1]. For this reason, one may expect that operators' networks are optimized to provide good indoor coverage and capacity for voice, video, and high-speed data services. However, surveys have shown that 45% of households and 30% of businesses experience poor indoor coverage [2]. This poor indoor coverage may lead to reduced subscriber loyalty and increased subscriber churn, which may significantly affect operators' revenues. As a consequence, vendors and operators are developing new solutions to address the indoor coverage problem.

A straightforward solution to enhance indoor coverage would be to increase the number of outdoor macrocell base stations (MBSs). Deploying a larger number of MBSs with a reduced cell radius may provide improved network coverage and capacity, but this approach is too expensive due to the high cost associated with MBSs. Moreover, this approach presents challenges in terms of site acquisition due to municipality and people's concerns about MBS towers [3]. It is also very difficult to achieve high indoor signal quality when providing coverage from outdoors due to wall attenuation losses. Therefore, providing indoor coverage from outdoors is not the best solution.

As a result, indoor solutions such as distributed antenna systems (DASs) and picocells have become attractive alternatives to provide services in indoor hotspots, e.g., shopping malls and office buildings. These operator-deployed solutions improve in-building coverage, enhance signal quality, offload traffic from outdoor MBSs, and allow high-data-rate services due to the fact that transmitters are closer to receivers.

Introduction

In the past decade, radio wave propagation modeling has attracted a great deal of interest from both academia and industry, because it facilitates efficient computation of path loss between a transmitter and a receiver in a given scenario, and plays an important role in, e.g., radio link planning and optimization processes.

Radio waves are electromagnetic waves, which can be decomposed into electric and magnetic fields. Along the propagation direction, these two fields are perpendicular to each other, creating the effect of polarization. Radio propagation is affected by the environment. For example, radio waves diffract on the edges of objects, reflections occur on an object when the wavelength is much smaller than the dimension of the object, scattering occurs if the object surface is not smooth, and attenuation losses depend on the material and size of an object.

The radio spectrum is the most precious resource in wireless communications. Different frequencies are typically allocated for use in different systems. The frequency of a radio wave also has a great impact on its propagation. A higher frequency yields a smaller wavelength and vice versa. On the one hand, radio waves with a higher frequency generally experience a higher attenuation loss, and thus propagate a shorter distance before the carried signal strength falls below a threshold. On the other hand, radio waves with a lower frequency have a larger wavelength, and can bypass obstructions more easily. Moreover, radio waves at adjacent frequencies may interfere with each other, hence it is necessary to carefully allocate frequency bands to avoid significant interference between different communication systems or even between radio links within the same system.

Driven by a new generation of wireless user equipments and the proliferation of bandwidth-intensive applications, mobile data traffic and network load are increasing in unexpected ways, and are straining current cellular networks to a breaking point. In this context, heterogeneous cellular networks, which are characterized by a large number of network nodes with different transmit power levels and radio frequency coverage areas, including macrocells, remote radio heads, microcells, picocells, femtocells and relay nodes, have attracted much momentum in the wireless industry and research community, and have also gained the attention of standardization bodies such as the 3rd Generation Partnership Project (3GPP) LTE/LTE-Advanced and the Institute of Electrical and Electronics Engineers (IEEE) Mobile Worldwide Interoperability for Microwave Access (WiMAX).

The impending worldwide deployments of heterogeneous cellular networks bring about not only opportunities but also challenges. Major technical challenges include the co-existence of various neighboring and/or overlapping cells, intercell interference and mobility management, backhaul provisioning, and self-organization that is crucial for efficient roll-outs of user-deployed low-power nodes. These challenges need to be addressed urgently to make the best out of heterogeneous cellular networks. This asks for a thorough revisit of contemporary wireless network technologies, such as network architecture and protocol designs, spectrum allocation strategies, call management mechanisms, etc. There is also an urgent need in the wireless industry, academia and even end-users to better understand the technical details and performance gains that heterogeneous cellular networks would make possible.

As discussed in Chapter 1, heterogeneous cellular networks (HCNs) with low-power nodes (LPNs) are important for improving the capacity and coverage of next generation broadband wireless communication systems. However, interference problems in HCNs pose an important challenge, and thus efficient interference management techniques are required to fully benefit from their deployments. The main contribution of this chapter is to review interference problems and interference management techniques for HCNs. A general notion of macrocell base stations (MBSs) and LPNs is adopted, but the simulations are based on Long Term Evolution (LTE) scenarios with macro eNBs and pico eNBs or femto HeNBs. More specifically, cell-selection and interference coordination methods are discussed, including mechanisms recently proposed in the 3rd Generation Partnership Project (3GPP) LTE, and their performances are evaluated through system-level simulations. In such simulations, LTE-specific notation is used, and macrocell user equipment (MUE) and picocell user equipment (PUE) denotes UEs served by macro eNBs and pico eNBs, respectively.

This chapter is organized as follows. First, Section 7.1 reviews the main reasons for excessive intercell interference in HCNs. In Section 7.2, due to its significance, range expansion (RE) for HCNs is treated in more detail. Some example simulation results that demonstrate the downlink (DL)/uplink (UL) coverage imbalance in heterogeneous deployments are also provided. Section 7.3 gives a high-level overview of intercell interference coordination (ICIC) methods that are applicable to HCNs, and the next three sections are dedicated to specific ICIC approaches: frequency-domain, power-based, and time-domain ICIC techniques are discussed in Section 7.4, Section 7.5, and Section 7.6, respectively.

Mobile broadband demands are increasing rapidly, driven by the popularity of various connected mobile devices with data services, such as smartphones, tablets, vehicles, machines and sensors. The notion of connected devices actually expands to encompass basically everything that can take benefits from a wireless connection. A true mobile broadband experience of high quality everywhere can be expected by consumers in the near future.

Mobile applications have become an indispensable part of people's everyday life, with requirements on seamless access to social media, video contents and cloud-based contents anytime, anywhere. To provide services that meet these requirements is of top priority for operators with ambitions to be a key wireless communications provider in the networked society. These requirements can only bemet by mobile networks with sufficient capacity and coverage. Mobile broadband today is mainly provided via networks based on UMTS Terrestrial Radio Access (UTRA) or Evolved UTRA (E-UTRA), and solutions differ in the details. Mobile networks need to evolve through improving the existing mobile broadband networks and adding more cells in an optimal way to migrate to a heterogeneous cellular network (HCN). The migration path could be different for different operators. A thorough understanding of the various components involved is vital for a cost-efficient, spectrum-efficient and energy-efficient network evolution.

This chapter provides an introduction to the whole book. First, the need for more capacity and mobile broadband forecasts are discussed in Section 1.1.

Introduction

Cognitive radio (CR) has recently become one of the most intensively studied paradigms in wireless communications. In its broadest sense, a CR can be thought of as an enhanced smart software defined radio (SDR). The terms SDR and CRwere introduced by J. Mitola in 1992 [1] and 1999 [2], respectively. SDR, sometimes shortened to software radio, is generally a multi-band radio that supports multiple air interfaces and protocols, and is reconfigurable through software running on a digital signal processor (DSP), field-programmable gate array (FPGA), or general-purpose microprocessor [3]. CR, usually built upon an SDR platform, is a context-aware intelligent radio capable of autonomous reconfiguration by learning from and adapting to the surrounding communication environment [4]. CRs are capable of perceiving and sensing their radio frequency (RF) environment, learning about their radio resources, user equipment (UE), and application requirements, and adapting their configuration and behavior accordingly. From this definition, two main characteristics of CR can be identified: cognitive capability (ability to capture information and learn from the radio environment) and reconfigurability (which enables the transmitter parameters to be dynamically programmed and modified according to the radio environment).

An important specific application often associated with CR is dynamic spectrum access (DSA). DSA, despite being a broader concept [5–7], is commonly understood as the reutilization of licensed RF bands by unlicensed UEs provided that the legitimate licensed UEs are not using the reused frequencies at a given time or in a given region of space.

Introduction

As the demand for high-rate wireless services increases, new techniques and architectures have emerged to increase their spectral efficiency and improve their reliability. During the last decade, multiple-input multiple-output (MIMO) or multiple-antenna technology has attracted much attention due to its ability to provide fast and reliable transmission without bandwidth expansion or increase in transmit power. For point-to-point MIMO systems, it has been shown that the capacity of an MIMO channel grows linearly with the minimum number of antennas at both ends [1]. For multi-user systems, MIMO can support space-division multiple access (SDMA) and provide multi-user diversity gain. MIMO has been a key to most modern wireless communication standards such as the 3rd Generation Partnership Project (3GPP) Long Term Evolution (LTE) and LTE-Advanced, Worldwide Interoperability for Microwave Access (WiMAX) and IEEE 802.11n Wireless Fidelity (WiFi).

While the original LTE mainly considered capacity, heterogeneous cellular networks (HCNs), where macrocells are overlaid with low-power nodes (LPNs) such as picocells, femtocells, and relay nodes, have attracted lots of interest in LTE-Advanced to meet the explosive but unequal mobile data traffic demands. On the other hand, due to the scarcity of spectrum, full frequency reuse has been an attractive strategy considered in LTE-Advanced [2]. In conventional homogeneous macrocell cellular networks operating under the principle of single-cell processing (SCP), there is strong intercell interference (ICI), which can be treated as noise and becomes the major challenge that limits system performance in terms of both throughput and fairness. In particular, user equipments (UEs) at the cell edges suffer the most from ICI.

Introduction

Cellular networks have been undebatably a success story, which resulted in wide proliferation and demand for ubiquitous heterogeneous broadband mobile wireless services. With the exponential increase in high-rate traffic driven by a new generation of wireless devices, data is expected to overwhelm cellular network capacity in the near future. Multi-tier heterogeneous cellular networks (HCNs) have been recently proposed as an efficient and cost-effective approach to provide unprecedented levels of network capacity and coverage. Cellular operators have started integrating small cells as a means to provide dedicated additional capacity either where most data usage generally occurs (i.e., enterprises, households) or where user equipments (UEs) are likely to experience poor data rate performance (i.e., cell edges, subway stations and households). Small cells such as femtocells offer radio coverage through a given wireless technology while a broadband wired link connects them to the backhaul network of a cellular operator. In conventional single-tier networks, the macrocell base stations (MBSs) have to cater to the needs of both outdoor and indoor UEs, which leads to poor indoor coverage and the appearance of dead spots [1–3]. In contrast, in femtocell-aided cellular networks, indoor UEs can enjoy high-quality wireless service from their designated femtocell access points (FAPs) in close proximity and outdoor UEs can experience higher capacity gains due to traffic offload by FAPs through the backhaul. Moreover, FAPs have the economical advantage of being less costly to manufacture and maintain as compared with MBSs.