Efficient Communication Protocols for Sensor Network Architecture with Multiple Mobile Sinks

Maximizing sensor network lifetime, efficiently gathering data from sensor nodes, and leveraging the reusability of sensor network protocols are crucial to make the emerging sensor network technology succeed in real world deployments. In this thesis, our aim is to develop novel communication protocols that will achieve this goal.;In the second part of this thesis, we turn our attention to leveraging our solution to support a wide class of applications with different data delivery requirements. We start by analyzing the limitation of the existing rendezvous data dissemination solutions. We then developed a novel adaptive rendezvous mechanism that efficiently adapts to the network irregularity as well as to the application query-requirements. We show how existing sinks cooperate as a single tier to build a reliable rendezvous structure. We emphasize synchronization of the underlying pull and push components to prevent resource wasting. The developed protocol works with equal efficiency in both regular as well as irregular sensor networks.;Finally, we propose a novel mobility-based protocol for structuring a generic infrastructure that could be efficiently leveraged by different upper layer protocols. The proposed protocol allows mobile robots/sinks to dynamically organize the arbitrary network topology into a physical co-centric circular layered infrastructure. The resulting infrastructure supports both multi-hop and data-mules regimes of communication and requires only local updates for maintenance. We prove the correctness and efficiency of the proposed scheme. We also provide a rough cost model that predicts the cost of communication over the resulting infrastructure.;First, a deployment-based solution of multiple mobile sinks is developed to maximize the lifetime of sensor networks. We introduced the concept of peeling phenomenon and show how sinks could move following the direction of the progressive peeling to retain the sink-to-network connectivity. We also proposed a novel data forwarding strategy, namely the "away-from-centroid" data forwarding, which increases the load balancing and leverages the utilization of the network resources. We then introduced the concept of a "guard region" that protects the core area of interest from being peeled-off and we show how this could contribute to further maximizing the network lifetime.