Enhancing Vehicular Anonymity in ITS: A New Scheme for Mix Zones and Their Placement

Abstract

Intelligent transportation systems (ITS) achieve improved throughput and safety using periodic vehicle-to-vehicle and vehicle-to-infrastructure wireless communication. Vehicles use pseudonyms which are frequently exchanged to protect against eavesdroppers using such messages for tracking. Such exchange takes place in mix zones where all wireless transmission is forbidden in order to prevent matching the new pseudonym of a vehicle with its previous one. Mix zones are not free: in addition to infrastructure cost, they impose a cost in terms of reduced vehicular throughput and disruption to vehicular communication. We present a scheme to manage traffic within a mix zone to make it more resilient to attacks against privacy. Second, we introduce a heuristic to place mix zones appropriately so that the gain in privacy is balanced against the cost in reduced throughput. We evaluate our schemes assuming a powerful attacker who has access to all wireless transmissions and uses a simple but powerful machine learning algorithm. Our algorithms are evaluated using detailed traffic simulations in two US cities: New York, NY, and Cambridge, MA.