]> 2006-06-05T10:00:00-05:00 2006-06-05T12:00:00-05:00 For this thesis we have built a system for having complex queries on physical environment, answered by the sensor network while addressing some of the main concerns such as energy efficiency, scalability, reliability and usability. The system takes inspiration from economical institutions such as corporations, factories, press and communities and a number of algorithms have been designed to implement the various processes of these institutions. The algorithms seek to organize the sensors into various types of organizations based on economic utility and they significantly reduce the system complexity and enable the building of large scale sensor networks. We have implemented the work and the simulation results shows that it has a much better performance in terms of the four factors stated in the previous paragraph, compared with the existing systems for Sensor databases.

The main contribution of this thesis is a new methodology for agent organization by extending the concept of cluster formation. This new methodology could be used for various distributed systems ranging from grid computing to robotic coordination. We validate our methodology by building this querying system for sensor networks and in the future we plan to propose to extend it for generic distributed systems.]]>