A data centric framework for mobile target tracking and data dissemination in wireless sensor networks

Wireless sensor networks are envisioned to be extremely useful for various environmental, civil, military and homeland security missions. Different from other paradigms of networks, sensor networks not only provide communication services but also generate sensing data to be transmitted. As sensor networks scale in size, so will the amount of sensing data generated. The large volume of sensing data coupled with the facts that the data are spread across the entire network, the sensor network is resource-constrained, and the sensor network is usually deployed in unattended and hostile environments, creates a demand for efficient and secure data collection and dissemination/access techniques to obtain useful and high-quality data from within the network.; In this thesis, we study the problem in a typical application scenario where a sensor network is deployed to monitor mobile targets in unattended environments and disseminate the obtained sensing data from dynamically changed detecting nodes to mobile sinks. We address the problem in three aspects.; Collaborative detection and efficient data collection. Since sensor nodes have limited sensing/communication ranges and are not reliable, it is necessary for multiple sensor nodes to collaborate in detecting a target. Further, due to the mobility of the targets, nodes involved in the collaboration will change over time. We propose a dynamic convoy tree-based collaboration (DCTC) framework to facilitate the collaborative detection among a dynamic set of sensor nodes, and efficient collection of data for generating high-quality sensing results. In DCTC, sensor nodes surrounding a mobile target form a tree structure, which can be dynamically adjusted, to facilitate the collaborations among them and collect their sensing data. One big challenge in implementing DCTC is how to reconfigure the convoy tree efficiently. We formalize the problem as an optimization problem of finding a convoy tree sequence with high tree coverage and low energy consumption, and propose both an ideal scheme and several practical schemes to solve it.; Efficient data dissemination. To facilitate users (static or mobile) in finding sensing data of interest from within a sensor network, we propose an index-based data dissemination scheme with adaptive ring-based index (ARI). This scheme is based on the idea that sensing data are collected, processed and stored at the nodes close to the detecting nodes, and the location information of these storing nodes is pushed to some index nodes. To tolerate failures and balance load, the index nodes for each type of targets form a ring surrounding the location which is determined based on the target type. To improve the performance, we further propose several mechanisms to optimize the ARI scheme.; Securing data forwarding. In data collection and dissemination, sensing data need to be forwarded in the network. To achieve authenticity and confidentiality in sensing data forwarding, innocent sensor nodes can share group keys for data encryption and authentication. However, the group key-based techniques will become ineffective if some nodes are compromised since the adversary may obtain group keys from these compromised nodes. To deal with node compromise, the innocent nodes should update their group keys to prevent the adversary from utilizing the captured keys. Because most previously proposed group rekeying schemes have high overhead and are not suitable for sensor networks, we design and evaluate a family of predistribution and local collaboration-based group rekeying (PCGR) schemes, based on the idea that future group keys can be preloaded to nodes before deployment, and neighbors can collaborate to protect and appropriately use the preloaded keys.