Survivability of Self-organizing Sensor Networks
by
David Erbach
Western Kentucky University
Coauthors: Kerry Haenftling, Indiana University Purdue University Fort Wayne

In many situations, it is desirable to maintain surveillance of an area, but it is not possible or desirable to do so with humans. Inexpensive autonomous sensors are a natural alternative. However, it may not be practical to place sensors in precise locations, or to replace nodes which fail. An example would be the frontier between two hostile countries.

In such situations, it is desirable to place very small sensors. Such sensors must be able to self-organize their communication network, and to restructure the network if a sensor fails for any reason. There is always a difficult design trade-off between cost (which should be low) and power, reliability, and duty cycle (which should be high.)

This talk will discuss the problem, and demonstrate a simulation developed by the second author.

Date received: May 1, 2009