There is an increasing interest of companies and government agenciesto snoop on people's daily lives the increasing difficulty for peopleto handle such scenarios. The need for private communications isperhaps greater than ever before. Officials at the NSA have stated that “if you have enough meta-data you don’t reallyneed content” and that “we kill people based onmeta-data”. People have long needed to keep the communications among themselves private, but, increasingly, they may want to conceal not only the messages that they exchange, but alsowith whom they are communicating---or even the fact that they are communicating at all. This latter type of communication is said to benot only confidential and anonymous but also deniable, in the sensethat despite exchanging messages, participants can plausibly deny thatany such exchanges ever took place.This dissertation develops techniques and systems that empower usersin physical proximity to have mechanisms for deniablecommunications. Our work builds from the observation of noise in the surrounding technologies like wireless networks or powerline networks. The thesis particularly uses noise instead of protocol obfuscation tocreate deniable channels between individuals who do not want any thirdparty to recognize that there is possible communication in progress.Working with collaborators at Georgia Tech, I have built two systemsto explore two approaches at the link layer of wireless channelin 802.11 protocol by means of Denali. Looking for alternate technologies Istumbled upon innocuous-lookingpowerline networksand led to the work Powerline Whisperer, where Iexplored using the physical layer in powerline cables to do deniablecommunication. Due to lack of available cover, the system does not presume any already established communication channel to exchange amessage but rather depends on the thermal noise and the electromagnetic interference due to devices present in the medium.
PDF
download
878987797
028-85220240
