As wireless networking increasingly dominates our means of communication, the need for privacy and security has gained prominence. Wireless networks, owing to the unprotected communication medium, are vulnerable to unauthorized access of networking information. For example, by merely observing transmission times of packets, a passive eavesdropper can decipher source-destination pairs and paths of data flow in a network. Unauthorized retrieval of such information, known as traffic analysis, is a violation of user privacy. It also provides crucial information for the jamming of network traffic and launching of a denial-of-service attack.
This research aims to establish an analytical framework for achieving anonymity in wireless networks. The key objectives are to establish a theoretical framework upon which provably anonymous and secure protocols for multiple access communication and anonymous networking can be developed and analyzed. Drawing ideas of anonymous mixing from Internet privacy, timing channel analysis from information theory, and intrusion detection from statistical inference of point processes, countermeasures against passive and active means of compromising anonymity are investigated. Parallel to the well known rate-secrecy trade-off in point-to-point communications, fundamental tradeoffs between anonymity and network performance metrics, such as throughput and delay, are investigated. Scheduling and routing protocols are developed to prevent unauthorized release of networking information to passive and active adversaries.
