Secure, distributed monitoring and reporting of data from wireless sensor networks.

摘要

Wireless sensor network (WSN) security research generally considers models based on abstract applications. Instead, I argue that WSN security solutions must be sensitive to the application and infrastructure. Specifically, I formulate a new notion of an application-specific security context as the combination of a potential attackers motivation and the WSN vulnerability. To reduce the vulnerability, WSN design must balance security with traditional objectives such as the cost, energy and efficiency to a degree proportional to the attackers motivation. I illustrate this argument via four example applications in an attempt to formulate a basis to derive more grounded and realistic assumptions for WSN security and develop cost-effective security solutions to handle application-specific vulnerabilities in WSNs.;I then delve quite deeply into the application-centric context of ubiquitous medical monitoring systems. Wireless sensor devices collect long term medical data that aid medical professionals in the care and treatment of the elderly and other patients with chronic diseases. Using wearable body sensor networks (BSNs), patients may live more mobile and independent lives than they would under direct medical observation. As a result, the quality of life can be improved and the cost for medical care can be reduced.;Although security issues in BSNs have been studied previously, most existing work focuses on monitoring patients in a fixed location such as a hospital, relying on centralized approaches that have poor scalability. I develop a new protocol to support secure and scalable anytime/anywhere monitoring of patients to give them full mobility, support security of sensitive medical data, and preserve location privacy of patients, while sharing sensor data collection and retrieval workloads among distributed storage nodes.;I then expanded the context to that of a patient who is being monitored and then experiences a medical emergency. In such a case, some variations must occur in the system to provide more rapid response, to ensure that the data has been read by emergency personnel, and to release the location information under this limited, but extremely necessary scenario. Here, I present an approach to provide these needed benefits while continuing to provide scalability, privacy, and data integrity.