Off-campus WSU users: To download campus access dissertations, please use the following link to log into our proxy server with your WSU access ID and password, then click the "Off-campus Download" button below.

Non-WSU users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Access Type

WSU Access

Date of Award

1-1-2010

Degree Type

Dissertation

Degree Name

Ph.D.

Department

Electrical and Computer Engineering

First Advisor

Loren J. Schwiebert

Second Advisor

Sandeep K. Gupta

Abstract

Wireless sensor networks are made possible by the continuing improvements in embedded sensor, VLSI, and wireless radio technologies. Currently, one of the important challenges in sensor networks is the design of a systematic network management framework that allows localized and collaborative resource control uniformly across all application services such as sensing, monitoring, tracking, data aggregation, and routing.

The research in wireless sensor networks is currently oriented toward a cross-layer network abstraction that supports appropriate fine or course grained resource controls for energy efficiency. In that regard, we have designed a unified role-based service paradigm for wireless sensor networks. We pursue this by first developing a Role-based Hierarchical Self-Organization (RBSHO) protocol that organizes a connected dominating set (CDS) of nodes called dominators. This is done by hierarchically selecting nodes that possess cumulatively high energy, connectivity, and sensing capabilities in their local neighborhood. The RBHSO protocol then assigns specific tasks such as sensing, coordination, and routing to appropriate dominators that end up playing a certain role in the network.

Roles, though abstract and implicit, expose role-specific resource controls by way of role assignment and scheduling. Based on this concept, we have designed a Unified Role-Assignment Framework (URAF) to model application services as roles played by local in-network sensor nodes with sensor capabilities used as rules for role identification. The URAF abstracts domain specific role attributes by three models: the role energy model, the role execution time model, and the role service utility model. The framework then generalizes resource management for services by providing abstractions for controlling the composition of a service in terms of roles, its assignment, reassignment, and scheduling. To the best of our knowledge, a generic role-based framework that provides a simple and unified network management solution for wireless sensor networks has not been proposed previously.