Differentially private queries in crowdsourced databases for net neutrality violations detection

Maria Silvia Abba Legnazzi, Cristina Rottondi, Giacomo Verticale
2017 2017 Ninth International Conference on Ubiquitous and Future Networks (ICUFN)  
Lawmakers and regulatory bodies around the world are asserting Network Neutrality as a fundamental property of broadband Internet access. Since neutrality implies a comparison between different users and different ISPs, this opens the question of how to measure net neutrality in a privacy-friendly manner. This work describes a system in which users convey throughput measurements for the different services they use to a crowdsourced database and submit queries testing their measurements against
more » ... he hypothesis of a neutral network. The usage of crowdsourced databases poses potential privacy problems, because users submit data that may possibly disclose information about their own habits. This leaves the door open to information leakages regarding the content of the measurement database. Randomized sampling and suppression of small clusters can provide a good tradeoff between usefulness of the system, in terms of precision and recall of discriminated users, and privacy, in terms of differential privacy.
doi:10.1109/icufn.2017.7993924 dblp:conf/icufn/LegnazziRV17 fatcat:qme47lbcbrhfpaaxe4cjbxrvwy