Issue No. 05 - May (2006 vol. 7)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/MDSO.2006.36
Lu Yan , Abo Akademi University
Distributed Sensor Networks
S. Sitharama Iyengar and Richard R. Brooks
Chapman & Hall/CRC 2004
Since the last decade, distributed sensor networks have been an active research area. The trend has been to move from a centralized, expensive, single-node platform to a set of distributed, cheap, and lightweight components that are collectively capable of far more complex tasks than any individual supernode. Wireless sensor networks got a kick-start at the turn of the new millennium, accelerated by the advancement of wireless networking and the maturing of mass production techniques. Clearly, wireless sensor networks will be a key to true ubiquitous computing: sensor networks will eventually permeate the physical world and provide grounding for the Internet.
In Distributed Sensor Networks, S. Sitharama Iyengar and Richard R. Brooks aim to provide a systematic introduction to wireless sensor networks. However, its organization is somewhat ad hoc.
Audience, scope, and organization
Distributed Sensor Networks is a textbook that in many ways started 10 years ago, when sensor networks were somewhat arcane. Iyengar and Brooks believed the only way to monitor the real world adequately was to use a network of devices. The breadth and depth of research needed to design these systems adequately were unknown, and little reference to sensor networks existed. To fill the gap, Iyengar and Brooks compiled this book of papers written from the 1990s on by acknowledged leaders in a broad range of technical sensor-network-design fields.
The book is logically organized into 10 sections, each containing several chapters from leading research institutions and independent experts. The first section is a brief overview of sensor networks, introducing what they are, their applications, and how they differ from traditional systems.
Section two discusses signal processing and sensor data interpretation issues, presenting topics such as signal diction, classification, and target tracking from different perspectives. Once signals and images have been locally processed, the next step concerns making global decisions from the local information, which this book usually calls "information and data fusion." Section three considers information-fusion technology at many levels and in many ways. This section also introduces typical data modalities and common applications.
Section four is dedicated to networking technologies. Two important issues are discussed: how to deploy the networks and how to maintain communication among nodes. These issues are interconnected, and both require monitoring node energy expenditures. This is a main concern in section five, which explains how power-saving and energy-harvesting techniques can maximize sensor nodes' lifetimes after deployment. Because sensor nodes are usually battery powered and energy constrained, power management is critical.
Once a sensor network is deployed in the real world, the needs for adaptation and self-configuration also become obvious. Wireless sensor networks consist of numerous unreliable nodes and communication links, which are subject to intermittent failures and degradations. There's a need for architectures, models, and algorithms geared toward an environment-adaptive and fault-tolerant network. Such a network could be approached by self-organizing, self-configuration techniques. Sections six and seven focus on these issues.
Section eight emphasizes deriving models and controllers for wireless sensor networks and elaborates a discrete-event-controller system for distributed surveillance networks. Section nine looks at specific sensor network implementations and lessons learned by field-testing the systems, and the final section discusses beamforming technology for wireless sensor networks.
This spring, I included Distributed Sensor Networks as course material for my graduate classes. Although the students liked it and used it frequently, nobody actually read the entire book, for several reasons. It's a bit lengthy for normal classes, and it lacks a clear thread that weaves the various topics together—a common problem for compilations with many contributors. A short introduction to the book's organization would greatly improve its readability. Additionally, some students complained about the inferior quality of figures and illustrations in some chapters; the authors should replace these images with higher-resolution ones.
As its title implies, Distributed Sensor Networks covers a broad range of technical issues of wireless sensor networks. With over 500 illustrations and 1,000 pages of in-depth information, this book serves as a good reference tool for interested researchers and graduate students. However, its approach isn't suitable for beginners. I recommend using this book as a reference rather than as a tutorial.