The Scope for and Risks to Integrating Meta-data Privacy into Networked Applications through “Add-on” Anonymous Communication Networks (ACNs)

Tariq Elahi, Florentin Rochet, Neenu Garg – University of Edinburgh

The aim will be to investigate how effectively and efficiently anonymous communication networks (ACN) can act as “add-ons” between applications that do not have privacy built-in by design (non-PbD apps) and the Internet at large.

We will investigate both connection-based low-latency applications (such as web-browsing) as well as message-based high-latency applications (such as network telemetry reporting or email). We will investigate the use of Tor for the former and state-of-the-art mixnets for the latter. For the low-latency applications, our approach will be to investigate novel and non-standard Tor client set ups, such as every non-PbD on a device accessing the Internet through a dedicated Tor process or all devices accessing the Internet through a single Tor gateway. We will adopt a similar approach for higher-latency applications.

We expect that this project will positively impact adoption and hence we also wish to investigate the extent for how these ACN would cope with the increased load as well as the schemes to prevent abuse by the larger user bases.

This project will displace the current folk wisdom that to add anonymity or privacy properties to a system the addition of sending all traffic through an ACN is sufficient. It will provide a nuanced approach to integrating anonymous networks (while we focus on Tor the results here can be extended to other PETs) to systems that are not originally privacy-aware. If successful, this project would provide the evidence (both of the analytic and empirical flavour) that spurs the integration of ACNs into systems as building blocks.