Dynamic clustering of multi-modal sensor networks in urban scenarios

The paper addresses the issue of self-adaptation of a multi-modal sensor network with mobile sensors to better observe and track events of interest in a changing urban scenario by presenting a software module (middleware) called Event-driven Network Controller (ENC) that resides at every sensor node in the network and is independent of the sensor type. ENC translates the requirements of the application layer into messages that are diffused locally with the purpose of clustering multi-modal sensor nodes in the vicinity of an event and dynamically changing the local network topology, all to enhance the quality of the multi-modal data fusion. ENC is implemented in NS-2 to show its applicability for tracking a mobile target in an urban scenario using a network of pressure, video, and magnetic sensors.