Browse Theses

Sensor networks are used in many realtime applications for collecting information from monitored environments and objects, such as moving vehicle tracking, battlefield reconnaissance, and habitat monitoring. Sensor networks are often deployed in applications that involve humans. Whenever people become actively involved, privacy concerns become more important. Protecting privacy means preventing private and sensitive information from being compromised.

Protecting privacy focuses on preventing both intentional and unintentional leakage of private information in public data. Before public data can be made publicly available, there need to be mechanisms that either obscure private information derivable from it or restrict access to it. Because of the open medium of wireless sensor networks, data transmitted in the network can easily be eavesdropped by adversaries. Data collected by sensor networks are often used by someone other than the people who participate in the sensor networks. All these properties of sensor networks make privacy concerns more urgent. In this thesis, we discuss the privacy issues in sensor network applications and propose several methods to mitigate the leakage of private information in sensor networks applications.

In sensor network applications, privacy includes both communication privacy and information privacy. Communication privacy means the privacy of both parties to communications in the network, such as protection of their identities. Information privacy means private information in the data being transmitted in the network, such as location information. In this thesis, we propose multiple methods to address both the communication privacy problem and the information privacy problem. Performance of these algorithms is analyzed through theoretical analysis and evaluated by experiments.