Role-based Hierarchical Self Organization Protocol for Ad hoc Wireless Sensor Networks

Efficiently self organizing a network hierarchy with specific assignment of roles (or tasks) to sensors based on their physical wireless connectivity and sensing characteristics is an important and challenging problem. In this paper, we extend the hierarchical connected dominating set (CDS) construction algorithm, proposed by Jie Wu, to develop our role-based hierarchical self organization algorithm for wireless sensor networks. The resulting self organized sensor network establishes a network-wide infrastructure consisting of a hierarchy of backbone nodes, and sensing zones that include sensor coordinators, and sensing collaborators (or sensing zone members). Our paper identifies the need for organizing a sensor network according to the tasks appropriate for each sensor node based on their initial deployment in the network. Past research in group-based (or hierarchical) sensor networks have ignored the possibility of utilizing both the physical communication and sensing characteristics to assign roles to sensor nodes. We demonstrate the effectiveness of our design, which considers both, through simulations.

Keywords: Sensor networks, self organization hierarchy, sensing coordinators, sensing zones, backbone nodes, sensing proximity value (SPV), cumulative sensing degree (CSD), and connected dominating set (CDS).

see full paper and WSNA'2003 presentation.