Protecting Individuals by Hiding Flow Information on Last-mile Links

Abstract

Last-mile links connect a core network to the end-user’s communication infrastructure. A single user’s traffic can flow alone at a link, and thus an adversary can eavesdrop on the communication at the link, deducing critical information. Even if the traffic is encrypted, an aggressive enemy may perform statistical analysis based on the detected traffic. The results of this analysis may yield information in the communication that is private, characteristics of the communications, such as destination node or utilized networking protocols. With this information individual’s privacy can be damaged. In this paper we introduce a system that enhances remarkably user privacy on last mile links by making it more difficult for an adversary to perform statistical analysis. The system is based on one of the onion routing realizations Tor that is improved with the Traffic Flow Confidentiality (TFC) technique. Either a network operator or a service operator can deliver this service to customers. Moreover, we have measured the performance of our system that conceals flow information on last-mile link and beyond enhancing individual’s privacy.