Multi-Criteria Decision Making in Sensor Networks

This decision making process for a mobile robot can be adapted to clusters of sensor nodes rather than an individual sensor node. Preference can be given to a particular cluster for management (power management, node density management, etc.) based on activity level, the number of sensor nodes, and the importance of the activity. If the number of nodes is critical for the given application, then it gets a high weighting factor and would require some nodes to be put to sleep. We change the behavior of the network by tuning one of the parameters (increasing sleep time) based on the needs of the application. We intelligently analyzed the characteristic of a deployed sensor network for a given application and adaptively changed its operational behavior to suit the changing demands. We showed that appropriate sensor action needs to be scheduled in order to achieve maximum utility.