| A wireless sensor network (WSN) is comprised of a few hundred or thousand autonomous sensor nodes spatially distributed over a particular region. Each sensor node is equipped with a wireless communication device, a small microprocessor, and a battery-powered energy source. Typically, the applications of WSNs such as habitat monitoring, fire detection, and military surveillance, require data collection, processing, and transmission among the sensor nodes. Due to their energy constraints and hostile environments, the main challenge in the research of WSN lies in prolonging the lifetime of WSNs.;In this dissertation, we present four different topology management protocols for K-coverage and load balancing to prolong the lifetime of WSNs.;First, we present a Randomly Ordered Activation and Layering (ROAL) protocol for K-coverage in a stationary WSN. The ROAL suggests a new model of layer coverage that can construct a K-covered WSN using the layer information received from its previously activated nodes in the sensing distance. Second, we enhance the fault tolerance of layer coverage through a Circulation-ROAL (C-ROAL) protocol. Using the layer number, the C-ROAL can activate each node in a round-robin fashion during a predefined period while conserving reconfiguration energy. Next, Mobility Resilient Coverage Control (MRCC) is presented to assure K-coverage in the presence of mobility, in which a more practical and reliable model for K-coverage with nodal mobility is introduced. Finally, we present a Multiple-Connected Dominating Set (MCDS) protocol that can balance the network traffic using an on-demand routing protocol. The MCDS protocol constructs and manages multiple backbone networks, each of which is constructed with a connected dominating set (CDS) to ensure a connected backbone network. We describe each protocol, and compare the performance of our protocols with Dynamic Source Routing (DSR) and/or existing K-coverage algorithms through extensive simulations.;The simulation results obtained by the ROAL protocol show that K-coverage can be guaranteed with more than 95% coverage ratio, and significantly extend network lifetime against a given WSN. We also observe that the C-ROAL protocol provides a better reconfiguration method, which consumes only less than 1% of the reconfiguration energy in the ROAL protocol, with a greatly reduced packet latency. The MRCC protocol, considering the mobility, achieves better coverage by 1.4% with 22% fewer active sensors than that of an existing coverage protocol for the mobility. The results on the MCDS protocol show that the energy depletion ratio of nodes is decreased consequently, while the network throughput is improved by 35%. |
