This paper studies the multi-agent average consensus problem under the requirement of differential privacy of the agents' initial states against an adversary that has access to all the messages. We first establish that a differentially private consensus algorithm cannot guarantee convergence of the agents' states to the exact average in distribution, which in turn implies the same impossibility for other stronger notions of convergence.
Submitted by Anonymous on Tue, 10/03/2017 - 12:08pm
Location: D'Agostino Hall, NYU School of Law
108 West 3rd Street | New York, NY, US
Contact:ccs@nyu.edu
Emerging technologies present intriguing challenges and exciting opportunities that require innovative thinking and diverse perspectives. Join us October 3rd as we take an interdisciplinary approach to exploring critical issues involving:
The 19th International Conference on Distributed Computing and Networking (ICDCN 2018), to be held between Jan 4-7, 2018 at IIT-BHU, Varanasi, India will have a workshop on Smart Cities: http://iitbhu.ac.in/ICDCN2018/
