Chapter 10
Sensor Deployment Abstraction
Simplicity and elegance are unpopular because they require hard work and discipline to achieve and education to be appreciated.
—Edsger Dijkstra
As we move into the future, the effects of Moore’s law will progessively make the unit cost of sensor devices more and more negligible. Indoor and even outdoor environments with access to the existing power grid will be fertile environments for large number of low-impact sensors. Similarly motor vehicles, which already use large number of microprocessors in their existing systems, will be able to easily evolve to include sensors and wireless communication devices as standard hardware. The effect of this will be a potentially information-dense environment, with over-whelming amounts of data available from anywhere inhabited by humans. In the long term, the greatest cost and the challenge involved in building sensor networks will rest not in the hardware, but in the software. In order to deal with large numbers of nodes measuring large amounts of data, whole new paradigms in data aggregation and network architecture will be needed. Networks on this scale will resemble a living system more than a series of traditional computer networks, and our approach to them must vary accordingly. The deployment of sensors often necessitates an understanding of certain requirements such as sensor coverage, redundancy, and network architecture providing the essence of sensing physical space. Applications such as ...
