During the past decade, increasingly advanced technologies have made it easier to compress, distribute, and store multimedia content. Multimedia standards, such as JPEG, MPEG, and H.26x, have been adopted internationally for various multimedia applications. Simultaneously, advances in wireless and networking technologies, along with a significant decrease in the cost for storage media, has led to the proliferation of multimedia data. This convergence of networking, computing, and multimedia technologies has collapsed the distance separating the ability to create content and the ability to consume content.

The alteration, repackaging, and redistribution of multimedia content pose a serious threat to both governmental security and commercial markets. The ability to securely and reliably exchange multimedia information is of strategic importance in fighting an unprecedented war against terrorism. A recent high-profile leak involved a classified video of Osama bin Laden's camp captured by an unmanned aerial surveillance vehicle, when one copy of the tapes shared between the Pentagon and CIA officials was leaked to the news media. Without effective traitor-tracing tools, different agencies would still be reluctant to share critical information, which jeopardizes the mission of fighting terrorism and defending national and global security. To prevent information from leaking out of an authorized circle, it is essential that the governments have the forensic capability to track and identify entities involved in unauthorized redistribution of multimedia information.

In addition to the demands from homeland security, preventing the leak of multimedia information is also crucial to the economy.

Until now we have discussed cooperation stimulation and how side information changes the behavior dynamics in various types of media-sharing social networks. For colluder social networks, as discussed in Chapters 5 and 8, before collusion, colluders need to reach an agreement regarding the fair distribution of the risk of being detected and the reward from illegal usage of multimedia. Chapters 5 and 8 analyze how colluders bargain with one another to achieve fairness of collusion, assuming all colluders report their private information (their received fingerprinted copies) honestly.

In reality, some colluders might break their fair-play agreement. They still wish to participate in and receive reward from collusion, but they do not want to take any risk of being detected by the digital rights enforcer. To achieve this goal, they may lie to other attackers about their fingerprinted copies. For example, they may process their fingerprinted signals before multiuser collusion and use the processed copies instead of the originally received ones during collusion. The cheating colluders' goal is to minimize their own risk while still receiving reward from collusion. Therefore, they select the most effective precollusion processing strategy to reduce their risk.

Precollusion processing reduces the cheating colluders' risk, and makes other attackers have a higher probability than the cheating colluders of being detected. It is obviously a selfish behavior. In some scenarios, precollusion processing can also increase other attackers' probability of being detected; this is not only selfish, but also malicious.

Chapter 9 studies cooperation stimulation for P2P video streaming over Internet and wireless networks. One assumption there is that all users in the P2P networks are rational, and their goal is to maximize their own payoffs. As discussed in Chapter 9 and shown in Figure 12.1, they may lie to others about their personal information if they believe cheating can help increase their utilities. There might also exist malicious users who aim to exhaust others' resources and attack the system. For example, in P2P systems, they can tamper the media files with the intention of making the content useless (the so-called pollution attack). They can also launch the denial of service (DoS) attack to exhaust other users' resources and make the system unavailable. What is more, once an attacker is detected, he or she can leave the network temporarily, come back later with a new ID, and continue causing damage to the system.

To further proliferate P2P systems and provide reliable service, misbehavior detection and attack resistance are fundamental requirements to stimulate user cooperation even under attacks. A challenging issue in malicious user detection in P2P video streaming is to differentiate between “intentional” misbehavior (for example, intentional modification of the video content) and “innocent” ones (such as transmission error and packet loss in error-prone and congested networks).

In this chapter, we first model the P2P video streaming network over the Internet as a multiplayer game, which includes both rational (selfish) and malicious users.

In the previous chapter, we used multimedia fingerprinting social network as an example and examined the impact of side information on users' strategies. We showed that information about the statistical means of detection statistics is useful to improve the detection performance. A straightforward question to ask is whether there is other information that may potentially influence user dynamics. In this chapter, we investigate how side information changes the risk–distortion relationship in linear video collusion attacks against Gaussian fingerprints.

Video data have the unique characteristic that temporally adjacent frames are similar but usually not identical. Video collusion attacks include not only the intercopy attack that combines the same frames from different copies, but also the intracopy attack that combines temporally adjacent frames within the same copy. Because temporally adjacent frames are not exactly the same, an intracopy collusion attack will introduce distortion. Therefore, for a video collusion attack, there exists a tradeoff between the fingerprint remaining in the colluded copy that determines colluders' probability of being detected – that is, their risk – and the quality of the colluded copy – the distortion. It is extremely important for colluders to learn the risk–distortion tradeoff, as knowing this tradeoff would help them choose the best strategy when generating the colluded copy. It is also essential for the fingerprint detector to understand the risk–distortion tradeoff, as it would help predict colluders' behavior and design an anticollusion strategy.

With recent advances in communications, networking, and computer technologies, we have witnessed the emergence of large-scale user-centered web 2.0 applications that facilitate interactive information sharing and user collaboration via Internet – for example, blogs; wikis; media-sharing websites such as Napster, Flickr, and YouTube; social networking services such as Facebook, LinkedIn, and Twitter; and many others. Different from traditional web applications that allow only passive information viewing, these web 2.0 sites offer a platform for users to actively participate in and contribute to the content/service provided. The resulting trend toward social learning and networking creates a technological revolution for industries, and brings new experience to users.

The emergence of these websites has significant social impact and has profoundly changed our daily life. Increasingly, people use the Internet as a social medium to interact with one another and expand their social circles, to share information and experiences, and to organize communities and activities. For example, YouTube is a popular video-sharing website on which users upload, share, and view a wide variety of user-generated video content. It targets ordinary people who have Internet access but who may not have a technical background on computers and networking, and enables them to upload short video clips that are viewable to the worldwide audience within a few minutes. Its simplicity of use and the large variety of content offered on the website attract more than one billion views per day, according to a blog by Chad Hurley (cofounder of YouTube) on October 9, 2009, and make video sharing an important part of the new Internet culture.

Different social networks may have different structures. The discussions in the previous chapters focused mainly on distributed scenarios. For example, there are no central authorities in the colluder social networks in Chapters 5 and 8, and the P2P systems in Chapters 9 and 12 are fully distributed, meaning that every peer takes the same role. In reality, some social networks have a centralized structure in which there are one or more entities whom all users trust and who can facilitate interaction among users. For example, the first generation P2P file-sharing networks (for example, the Napster music file-sharing system) used a set of central servers to provide content indexing and search services. Although these servers cannot enforce user cooperation, they can facilitate user interaction. Other media-sharing social networks have a distributed structure and a flat topology in which users take the same role – for example, Gnutella and Chord. Distributed schemes should be designed for such social networks. In this chapter, we use colluder social networks in multimedia fingerprinting as an example to investigate the impact of network structure on social networks.

In colluder social networks, as discussed in Chapters 5 and 8, colluders aim to achieve fair collusion, which requires all colluders to report their received fingerprinted copies honestly. As discussed in Chapter 11, the assumption of fair play may not always hold, and some colluders may process their fingerprinted copies before collusion to further lower their risk of being detected.

As introduced in Chapter 2, multimedia fingerprinting systems involve many users with different objectives, and users influence one another's performance. To demonstrate how to analyze human behavior, we take the equal-risk fairness collusion as an example.

During collusion, colluders contribute their fingerprinted copies and collectively generate the colluded copy. As demonstrated in Section 2.3.2, depending on the way attackers collude, different colluders may have different probabilities to be detected by the digital rights enforcer. Each colluder prefers the collusion strategy that favors him or her the most, and they must agree on risk distribution before collusion. A straightforward solution is to let all colluders have the same probability of being detected, which we call equal-risk fairness. Depending on the fingerprinting system and the multimedia content, achieving equal-risk collusion may be complicated, especially when colluders receive fingerprinted copies of different resolutions, as shown in Section 2.3.2. In this chapter, we use equal-risk fairness as an example, analyze how colluders select the collusion parameters to achieve fairness in scalable video fingerprinting systems, and provide a simple example on behavior dynamics analysis and the investigation of the impact of human factors on multimedia systems.

In this chapter, we investigate the ways in which colluders distribute the risk evenly among themselves and achieve fairness of collusion when they receive copies of different resolutions as a result of network and device heterogeneity. We also analyze the effectiveness of such fair collusion in defeating the fingerprinting systems.

In the previous chapter, we used colluders in multimedia fingerprinting as an example to study the necessary conditions for colluders to cooperate with one another, and investigated how attackers negotiate with one another and reach an agreement. In this chapter, we consider P2P networks, investigate how users in P2P systems cooperate with one another to form a social network, and study the optimal cooperation strategies.

As introduced in Chapter 3, mesh-pull–based P2P video streaming is one of the largest types of multimedia social networks on the Internet and has enjoyed many successful deployments. However, because of the voluntary participation nature and limited resources, users' full cooperation cannot be guaranteed. In addition, users in P2P live streaming systems are strategic and rational, and they are likely to manipulate any incentive systems (for example, by cheating) to maximize their payoffs. As such, in this chapter, we study the incentives for users in video streaming systems to collaborate with one another and design the optimal cooperation strategies.

Furthermore, with recent developments in wireless communication and networking technologies and the popularity of powerful mobile devices, low-cost and high-quality–service wireless local area networks (WLANs) are becoming rapidly ingrained in our daily lives via public hotspots, access points, digital home networks, and many others. Users in the same WLAN form a wireless social network; such wireless social networks have many unique properties that make cooperation stimulation more challenging.

In the previous chapter, using colluder social networks in multimedia fingerprinting as an example, we showed that the network structure can affect misbehavior detection and and the overall system performance. In this chapter, we investigate the impact of network structure on the optimal cooperation strategies in hybrid P2P streaming networks, in which some users with very high interconnection bandwidth act jointly as one user to interact with the rest of the peers.

Although P2P video streaming systems have achieved promising results, they also introduce a large number of unnecessary traverse links, which consequently leads to substantial network inefficiency. However, in reality, every peer can have a large number of geographically neighboring peers with large intragroup upload and download bandwidth, such as peers in the same lab, building, or campus. If these peers have special collaboration among themselves and work jointly as one user toward the rest of the network, the unnecessary traverse links can be reduced. In this chapter, we denote those geographically neighboring peers with large intragroup upload and download bandwidths as group peers. To reduce the unnecessary traverse links and improve network efficiency, instead of considering each peer's strategy independently, we investigate possible cooperation among group peers and study their optimal collaboration strategy.

Because of the heterogeneous network structures, different group peers might take different roles during cooperation. In this chapter, we investigate the optimal cooperation strategies under different compositions of group peers, including the scenarios with or without central authorities, and examine whether peers can have different amounts of resources.

Medium access control (MAC) is of paramount importance in wireless systems: it orchestrates how the spectrum is shared across users and flows directly impacting the system throughput, reliability, quality of service (QoS), and fairness. Numerous works in the literature have challenged the classical layered view of protocol stacks in order to improve the poor utilization of the scarce spectrum resources [1]. Both in the context of random access and contention-free access, substantial gains have been demonstrated by making the MAC layer more aware of channel conditions and applications. Another classical view that has been challenged over the years is the tendency to emulate a point-to-point link view over inherently point-to-multipoint wireless medium. In the classical approach, packets received by unintended users are simply discarded. Originally, in the multihop routing domain, and more recently in the single-hop case, the notion of unintended user has become stale, especially in the contexts of cooperative communication, network coding, and opportunistic routing [2–5, 22].

In this chapter we focus primarily on a specific network scenario, where there is only one wireless transmitter serving many receivers. We assume a contention-free MAC: a centralized scheduler dynamically allocates channels (e.g., spreading codes and frequency subbands) to multiple users over time.

This chapter first presents an overview of the possible signaling techniques in relay networks. The differences between the formation of the signals of the relays, network performance, and complexity are highlighted. Then, the problem of relay selection is reviewed and key factors in the design of relay selection schemes are described. Finally, as a case study, the details of one exemplary relay selection protocol are presented.

Introduction

The demand of high data rates due to the explosive growth in data centric applications drives further innovations in wireless communication systems. Novel techniques have been developed to improve the reliability of wireless links and to boost their data rates. The main novelty in those techniques in the last decade is to exploit the characteristics of the wireless medium rather than to suppress its features. Examples of these techniques include opportunistic communications, multiple-input-multiple-output (MIMO) communications, and cooperative communications (see reference [1] and references therein). Cooperative communications exploit one of the main features of the unguided wireless medium: the broadcast feature. The broadcast feature has been regarded as a negative feature, since it is the source of the interference dilemma of radio communications. The same feature allows nearby nodes of a transmitting source to overhear the transmission and in turn to possibly relay the signal to a destination. Thus, nodes can cooperate to overcome the limitations set by the wireless channel. This idea has led to the notion of cooperative communications [2].

This chapter discusses the impacts of channel estimation on the reliability of ultrawideband (UWB) systems when path delays and path amplitudes are jointly estimated [1]. The Cramér–Rao bound (CRB) for the path delay estimates is presented as a function of the signal-to-noise ratio (SNR) and signal bandwidth. The performance of a UWB system employing a Rake receiver and maximal ratio combining (MRC) is analyzed taking into account estimation errors as predicted by the CRB. Expressions for the bit error rate (BER) are obtained displaying the effects of the number of pilot symbols and the number of multipath components on the overall system performance. Transceiver design issues, such as allocation of power resources to pilot symbols, signal bandwidth, and the number of diversity paths (fingers) used at the receiver, are discussed in the context of the effects of estimation errors. Allocations of power resources to pilot symbols are determined to optimize the BER. Finally, the estimation errors are taken into account to optimize the signal bandwidth and the number of fingers of the Rake receiver in UWB systems.

Introduction

One of the most attractive features of UWB is its ability to resolve multipath. Numerous investigations have confirmed that the UWB channel can be resolved into a significant number of distinct multipath components [2–4]. A Rake receiver with MRC can be employed in UWB systems to exploit the multipath diversity. However, Rake receivers require knowledge of multipath delays and amplitudes.

Wireless systems are commonly affected by interference from various sources. For example, a number of users that operate in the same wireless network can result in multiple-access interference (MAI). In addition, for ultrawideband (UWB) systems, which operate at very low power spectral densities, strong narrowband interference (NBI) can have significant effects on the communications reliability. Therefore, interference mitigation and awareness are crucial in order to realize reliable communications systems. In this chapter, pulse-based UWB systems are considered, and the mitigation of MAI is investigated first. Then, NBI avoidance and cancelation are studied for UWB systems. Finally, interference awareness is discussed for short-rate communications, next-generation wireless networks, and cognitive radios.

Mitigation of multiple-access interference (MAI)

In an impulse radio ultrawideband (IR-UWB) communications system, pulses with very short durations, commonly less than one nanosecond, are transmitted with a low-duty cycle, and information is carried by the positions or the polarities of pulses [1–5]. Each pulse resides in an interval called “frame”, and the positions of pulses within frames are determined according to time-hopping (TH) sequences specific to each user. The low-duty cycle structure together with TH sequences provide a multiple-access capability for IR-UWB systems [6].

Although IR-UWB systems can theoretically accommodate a large number of users in a multiple-access environment [2, 4], advanced signal processing techniques are necessary in practice in order to mitigate the effects of interfering users on the detection of information symbols efficiently [6].