Dynamic Deployment of Mobile and Stationary Sensor Nodes for the Rapid Exploration of Emergency Areas

Abstract : The study of distributed event processing is of major importance in security and surveillance applications. The project is envisioned to assist mission-critical operations by providing command and control systems, which collect sensor information and react promptly to the detection of suspicious high-level events. Events are monitored, composed and propagated within a wireless sensor network that consists of nodes with varying sensing capabilities, including cameras, microphones, magnetometers, temperature and infrared sensors. Detecting complex events in a sensor network includes 1) selecting a suitable event execution plan, i.e. event operators (e.g. event automata), and their order of execution, and 2) carefully placing event operators to physical nodes in the network. This effort developed three novel search algorithms suitable for a distributed control environment, Multiple Depth First Search (MDFS), Brick and Mortar, and Hybrid Exploration. They are effective in environments with the following assumptions: No a priori knowledge, lack of GPS positioning, and no centralized control of agents. The Hybrid Exploration algorithm does make use of RFlD tags and use 5 tag-to-tag communications, and assumes multi-hop communications as in a deployed sensor network.