Orthogonal frequency division multiplexing (OFDM)was proposed in the 1960s (see Chang and Gibbey (1968)) and has been actively investigated since then. It can be used for both wired and wireless communications, providing several attractive features. One important feature of OFDM is that it is ISI-free. In OFDM, data symbols are transmitted by multiple orthogonal subcarriers. Each signal transmitted by a subcarrier has a narrow bandwidth and experiences flat fading without interfering with the other subcarriers' signals. From this, a simple one-tap equalizer can be used in the frequency domain to compensate for fading, while a complicated equalizer is required in a single-carrier system to overcome ISI.

It is generally known that OFDM will not outperform single-carrier systems (in terms of the average BER) when a single modulation scheme is used for all subcarriers. However, OFDM can offer a better performance if adaptive bit loading is employed. Since each subcarrier may experience different fading, the SNR varies among the subcarriers. A different number of bits per symbol can be transmitted using a different modulation scheme across subcarriers depending on the SNR for each subcarrier. For example, subcarriers with low SNR may transmit no signal or may use a lower-order modulation to stay below a certain BER ceiling, while more bits per symbol can be transmitted through subcarriers with high SNR. This approach of adaptive bit loading is used for wired communication systems (Bingham, 1990). Indeed, adaptive bit loading allows OFDM to outperform single-carrier systems. However, in some wireless communication systems, including digital terrestrial TV broadcasting, adaptive bit loading becomes impractical to implement in compensating for different fading across subcarriers.

This chapter's main objective is to discuss and to some extent dispel some common myths and misconceptions associated with interference mitigation solutions. Our goal is to shed some light on the lessons learned while researching and developing solutions.

A common path taken in the development of interference mitigation techniques often begins by identifying solutions developed for different purposes and applying to the problem at hand. In general, this path constitutes an extremely powerful approach, and examples given in Chapter 7, including time and spectral multiplexing, clearly demonstrate the effectiveness of the resulting solutions. However, applying solutions out of the original context for which they were developed is no simple task, since it requires a careful examination of all the parameters and the assumptions that come into play. It is often when this step is overlooked that myths are constructed.

Contrary to common belief, we show that some techniques often associated with interference mitigation do not constitute solutions. These techniques may in fact aggrevate the interference problem or have a negative impact on the overall system performance. They constitute what we call pitfalls that should be avoided if possible.

We find two recurring myths in most pitfalls studied, although this list is far from exhaustive.

1. Dealing with interference is similar to dealing with random noise and other wireless channel propagation properties and impairments.

2. A set of system parameters such as transmitted power, offered load, packet size, error correction scheme, and modulation techniques can be optimized in order to mitigate interference.

Our objectives in this chapter are to describe the basic building blocks in performance evaluation as we focus on identifying and understanding the effects of interference in wireless communications and its impact on system performance.

Since we set out to evaluate the effects of interference on performance, the first question we ask is what is interference? The term “interference” has been extensively used in the context of communication, in both wired and wireless systems. While an accurate definition may be dependent on the specifics of the context considered, the term generally refers to signal impairments due to factors in the environment such as channel propagation properties, other radiated power, and noise.

The second question is concerned with the performance evaluation of interference, namely, what are the quantitative measures that characterize interference, and consequently how should the resulting level of performance be quantified? One interference metric that has been used extensively includes the so-called signal to interference ratio. However, this measure does not characterize completely the resulting performance since performance is often tied to the quality of service requirements, which vary depending on the application considered. Our objective is to provide a list of performance metrics that can accurately quantify the network performance from an application perspective.

Since not all systems behave in the same way given the same level of interference, an important aspect of performance evaluation is to identify parameters that impact performance.

Wireless networks are rapidly becoming a part of the ubiquitous computing environment, and whether they are enterprise networks or in public hot spots (for example in airports, hotels, homes), often they are deployed in infrastructureless environments. The rapid specification development phase and the tight time to market cycle that follows leave little room for performance enhancements and proper coexistence consideration.

Why did I write this book?

Having gone through a somewhat complete performance analysis and coexistence development cycle for wireless network technologies being developed by the IEEE 802 standard working groups, and having gained some experience on the topic, I feel compelled to share it with other network engineers and researchers that are pursuing similar objectives. In particular, I would like to share the methodologies developed and the lessons learned from this process with others embarking on a similar quest.

The audience for this book includes: (1) researchers interested in performance evaluation and interference mitigation techniques; (2) wireless systems engineers and practitioners designing wireless communication systems; (3) users of wireless networks.

This book is unique because it focuses on a system level view of the problem of interference and its solution space. Generally, interference is dealt with at the physical layer. There are several outstanding books that focus on the accurate characterization of the wireless channel in addition to the development of physical layer techniques for filtering and anti-jamming.

The main themes of this book are to explore evaluation methods for quantifying the mutual effects of interference on the performance of wireless networks and to investigate system-level solutions for their coexistence in the same environment.

The coexistence of wireless communication systems operating in the same environment has become a “hot” topic in recent years as more systems are choosing to use the unlicensed bands and forfeiting the need to purchase spectrum.

There are two specified unlicensed bands for the operation of wireless systems, namely:

1. the industrial scientific and medical (ISM) band that includes the 900 MHz, 2.4 GHz, and 5.8 GHz frequencies;

2. the unlicensed national information infrastructure (UNII) band that includes the 5.2 GHz band. This band was opened in 1997 in the United States in order to expand broadband access opportunities.

Few rules apply in the unlicensed bands such as the ISM band. For example, the rules defined in the Federal Communications Commission Title 47 of the Code for Federal Regulations Part 15 relate to the total radiated power and the use of the spread spectrum and frequency hopping modulations. It is commonly understood that all users of the unlicensed bands can equally affect the quality and the usefulness of this spectrum. Thus, the major downside of the unlicensed band is that frequencies must be shared and potential interference tolerated.

We distinguish between several types of users in these unlicensed bands.

This chapter is designed to give the reader a comprehensive understanding of the fundamentals in wireless protocol design. First, we overview some of the physical layer and the medium access control layer design choices. Then, we give the details of select major protocols as examples of the concepts described.

Physical layer

The physical layer has the main function of transporting the information bits passed by the higher layers over a physical medium and recovering them on the other side of the medium. We can view the physical layer in terms of a digital or analog communication channel and modules that map digital information to an analog signal in case the channel is analog. Figure 2.1 illustrates the main components of the physical layer that are discussed in the following sections. For an in-depth treatment of communication systems, the reader is referred to other texts.

Communication channel

A communication channel consists of a physical medium, such as radio waves, copper wire, optical fiber, and the associated equipment necessary to transmit information over the medium. Communication channels can be used for either digital or analog transmission. Digital transmission consists of transmitting a sequence of pulses corresponding to a sequence of information bits. Analog transmission involves the transmission of waveforms associated with the transmitted signal. The bandwidth of a channel, W, measures the width of the window of frequencies that are passed by the channel.

A closed-loop modeling environment captures the mutual effects of interference on each end device including the protocol interactions. It generally consists of detailed simulation models for the MAC and PHY layers of the wireless technologies under consideration. Additionally, a major part of any network protocol simulation model is the accurate characterization of the application and the general configuration considered. This is commonly known as usage models that describe a user activity and the deployment environment.

Detailed simulation models vary in size and complexity. Given CPU and memory requirements to make use of detailed models, it is not uncommon to find detailed protocol simulations combined with mathematical models to approximate some parts of the system, most typically the channel and physical layer models.

In this chapter, we first describe what constitutes usage cases and overview the major components that define them. In particular, we overview application models, network topologies, and channel propagation models. Then we focus on the modeling of network protocols, including the MAC and PHY layers. We discuss how to speed up this model by using some mathematical approximations for the channel and PHY models. An example using the simulation modeling concepts is presented in the context of a case study for assessing the interference between IEEE 802.11b and Bluetooth. Finally, the simulation results obtained from the closed-loop model are compared with the open-loop model results described in Chapter 4.

The main goal of this chapter is to describe proven techniques employed to mitigate interference. We focus on dynamic and system level mechanisms that are able to adapt to the interference environment. Interference suppression techniques including coding, modulation, and filtering, in addition to others related to physical layer technologies such as CDMA and OFDM, abound in the open literature. The interested reader is referred to other texts such as ref. for an in-depth treatment of these technologies.

The system level coexistence techniques that we are concerned with can be classified into two broad categories. The first category of solutions consists of some form of sharing, making use of either temporal or spectral sharing, and, in some instances, a joint time and frequency domain technique. The second category of solutions is about adaptation and the opportunity to choose either the radio or the network that is best suited to the environment. This class of solutions includes handovers and the ability to roam across different networks.

Sharing the medium is synonymous with multiple access techniques. In Chapter 2, we briefly overviewed three types of multiple access techniques, namely TDMA, FDMA, and CDMA. Since the emphasis is on system level solutions, we consider TDMA and FDMA. Although CDMA is considered to be an effective interference suppression technique, it is specific to coding functionality at the physical layer and therefore it is not pertinent to a system level solution.

An accurate assessment of channel conditions represents a first step in any interference mitigation strategy. This channel assessment is generally performed at the receiver and used by the transmitter in order to make an informed decision about the channel state. In some cases, the channel assessment is also performed at the transmitter's side. This is common in most carrier sense multiple access systems, where devices have to “listen” to the medium before transmission. In this case, the transmitter and the receiver devices are assumed to be closely located and therefore have the same channel conditions. However, it is not uncommon for the receiver side to be experiencing different channel conditions from the transmitter. This situation is also known as the hidden node problem and conversely the exposed node problem in carrier sense multiple access sytems. Therefore, to optimize communication it is critical that each transmitter and receiver pair maintains the state of the channel as seen by the receiver.

Given this need to assess the channel conditions per transmitter and receiver pair, there are two basic channel estimation strategies. Channel estimation can be based on either explicit or implicit methods. Explicit methods include bit error rate calculation, packet loss, or frame error rate measurements performed on each receiver. The measurements are then conveyed to the transmitter device at regular time intervals. Alternatively, implicit methods do not require the transmitter and receiver devices to exchange information about the state of the channel.

Modeling interference is difficult since it requires considering simultaneously the interactions between the interferer and the victim system. This is known as closed-loop performance modeling and will be discussed in greater detail in Chapter 4. In this chapter, we discuss several approximations to evaluating interference at the victim's receiver while ignoring the interactions between the interferer and the victim system. This open-loop evaluation is generally a lot easier to model.

The first model we discuss deals with approximating the interference with white noise and deriving a probability of bit error at the receiver. This theoretical bit error calculation depends on the modulation, the energy of the signal transmitted, and the noise or interference levels. While not feasible in real implementations, most receiver designs include a probability of bit error calculation in the function of the signal to noise ratio. The waterfall shape of the bit error curve helps determine the optimal operating point for a particular design.

The second model is concerned with an n-state Markov model for characterizing the state of the wireless channel. This technique goes back to early work by Gilbert that models the wireless channel using a two-state Markov chain where one state corresponds to a noisy channel and the other state corresponds to a noise free channel. We show how this model can be modified in order to model an interference limited and an interference free channel.

Secret-key distillation (SKD) is the technique used to convert some given random variables, shared among the legitimate parties and a potential eavesdropper Eve, into a secret key. Secret-key distillation is generally described as a protocol between the legitimate parties, who exchange messages over the public classical authenticated channel as a function of their key elements X and Y, aiming to agree on a common secret key K.

We assume here that the eavesdropper's knowledge can be modeled as a classical random variable. For more general assumptions on the eavesdropping techniques, please refer to Chapter 12.

The quantum transmission is assumed to be from Alice to Bob. In order to be able to specify another direction for secret-key distillation, we use the Claude-and-Dominique naming convention. Here, the random variables X and Y model the variables obtained by Claude and Dominique, respectively, using the quantum channel, and the random variable Z contains anything that Eve was able to infer from eavesdropping on the quantum channel. As detailed in Section 11.3.2, both assignments (Alice = Claude ∧ Bob = Dominique) and (Alice = Dominique ∧ Bob = Claude) are useful.

In this chapter, I first propose a general description of the reconciliation and privacy amplification approach. I then list the different characteristics that a SKD protocol can have. And finally, I overview several important classes of known results and treat the specific case of SKD with continuous variables.