Self-organization is a rather fascinating concept that enables systems consisting of huge numbers of autonomously acting subsystems to perform a collective task. Moreover, selforganizing systems show an overall behavior that cannot easily be predicted or even preprogrammed in a scalable way. It was in the early 1960s that people like Ashby and Eigen investigated self-organization properties in (natural) systems. Since then, a great number of (technical) solutions have been developed, which, either on purpose or unintentionally, inherently formed the basic concepts of self-organization.
The aim of this book is to investigate the concepts of self-organization in the context of autonomous sensor and actor networks. The primary objective is to categorize the basic self-organization methods and to survey techniques for communication and coordination in massively distributed systems according to the developed classification scheme. Basically, two possible approaches can be thought of for organizing this book. First, we could start analyzing sensor and actor network technology and figure out what basic mechanisms are employed and how these relate to self-organization. A second approach would be to introduce self-organization as a methodology, apparently used everywhere in our life (in nature and in technical systems), and afterwards to continue with technical issues in sensor and actor networks, searching for previously learned self-organization methods. I decided to follow the second approach in order to keep the focus on self-organization while studying the term in the world of sensor and actor networks. The term ‘self-organization’ is still often misunderstood and misinterpreted. Therefore, this textbook is intended to be a basis for a better understanding of the concepts of self-organization, especially in the domain of sensor and actor networks. It provides a stepwise introduction of definitions, methodologies and corresponding techniques relevant in the context of self-organization.
Recent advances in miniaturization and wireless communication enabled the development of low-cost sensor nodes. Additionally, new application domains of sensor and actor networks emerged that demand huge numbers of interacting devices. Thus, the relevance of self-organization methods is rapidly increasing, as it is considered the primary control paradigm for distributed and massively distributed systems. The reader will see that selforganization has a number of advantages compared with other control paradigms. So, it becomes possible to operate huge numbers of collaborating subsystems, even in cases of limited resources, unreliable communication and massive failures of single systems. Unfortunately, these advantages are accompanied by some rather annoying side effects, such as the increasing complexity and a nondeterministic behavior. By using optimal combinations of the basic methods of self-organization, these disadvantages can be minimized to some extent.
According to the objective of this textbook – to study sensor and actor networks – the most relevant domains of communication and coordination are deeply investigated based on well known algorithms and mechanisms and a number of case studies. This includes networking aspects of medium-access control, ad hoc routing, data-centric communication and clustering techniques. Additionally, control mechanisms for cooperation, task and resource allocation, and collaborative actuation are investigated. The book is concluded by a brief introduction to the domain of bio-inspired algorithms. This study is included for two reasons – first, to demystify the term bio-inspired networking, and, secondly, to show the capabilities of such bio-inspired approaches.
What is unique about this textbook?
This book represents the first comprehensive overview of self-organization techniques in the context of wireless sensor and actor networks. It also provides a detailed classification of the basic mechanisms of self-organization. There are many reasons to study self-organization, such as the fascinating effects, which, if correctly understood and employed, provide the possibility of envisioning new kinds of previous system limitations. Additionally, this book is the first comprehensive study of technical solutions focusing particularly on sensor and actor networks.
This textbook is intended for graduate students, researchers and practitioners who are interested in the broad field of self-organization techniques as well as in application domains in sensor and actor networks. The book is structured to accompany a graduate course in computer science. Thus, some basic knowledge of networking, communication protocols and distributed systems is required. As this textbook provides a global view of algorithms and protocols developed for building self-organizing networking architectures, it can also serve as a reference resource for researchers, engineers and developers working in the field of sensor and actor networks.
Structure and organization
The book is organized into five parts. We start with an introduction to self-organization as a control paradigm for massively distributed systems. Thus, Part I introduces the main ideas and concepts of self-organization. It can be seen as a reference for general studies in the field of self-organization. All relevant terms are introduced, based on examples of natural and technical self-organization. The primary intention is to become familiar with concepts of self-organization and to understand its opportunities and limitations.
Networking aspects relevant to ad hoc and sensor networks are investigated in Part II. This part can be regarded as a broad introduction to algorithms and protocols needed to develop and to maintain wireless sensor networks. Therefore, aspects of protocols for wireless communication, ad hoc routing, data-centric communication and clustering are discussed. The bridge to self-organization methods introduced in Part I is built in the form of permanent discussions of self-organizing aspects of investigated algorithms.
Coordination and control aspects are studied in Part III. This part starts with an introduction to the concepts, challenges and opportunities that emerged with the development of sensor and actor networks. Besides the networking issues discussed in Part II, communication and coordination are relevant for sensor–actor control. Additionally, concepts of task and resource allocation are investigated that allow collaborative executions of complex tasks by autonomously self-organizing systems. Again, the basic self-organization methods are outlined in all analyzed techniques.
Part IV is intended as a conclusion and summary of the investigations in the previous parts of the book. The basic self-organization methods are revisited in the context of the algorithms, techniques and protocols investigated in the context of sensor and actor networks. Additionally, evaluation criteria are discussed that are relevant for estimating the quality and performance of self-organization techniques in sensor and actor networks.
Finally, Part V introduces a very special field that is strongly related to self-organization, namely bio-inspired networking. After a brief introduction to this research domain, the principles and concepts of three selected areas of bio-inspired research are investigated. In all three domains, case studies are depicted that provide solutions for the efficient operation of sensor and actor networks.
Since I began working on this textbook on self-organization in sensor and actor networks, many people have given me invaluable help and have been influential in shaping my thoughts on how to organize and teach a course on this topic. I want to thank all these people, including my students, colleagues, faculty members and helpful friends from around the world. Namely, I would like to mention Ian F. Akyildiz and Adam Wolisz, who encouraged me to stay with this topic, Imrich Chlamtac, who shared my visions in the bio-inspired networking domain, Özgür B. Akan, who helped me in various ways and with whom I coedited a special issue on bio-inspired computing and communication for Ad Hoc Networks, Reinhard German, who encouraged me to initiate my studies on integrated robot-sensor networks, and to Bettina Kr¨uger, Isabel Dietrich and Christoph Sommer for proofreading the manuscript. Last, but not least, I wish to thank Birgit Gruber from Wiley, who contacted me at an Infocom conference and convinced me of the idea to write a book on this topic, and all the staff at Wiley (Sarah Hinton, Wendy Hunter, Richard Davies and Joanna Tootill) for their assistance during the writing and the production phase of this book.
Falko Dressler - Erlangen, Germany