Quasi-Linear Cryptanalysis of a Secure RFID Ultralightweight Authentication Protocol

Abstract

In 2010, Yeh, Lo and Winata [1] proposed a process-oriented ultralightweight RFID authentication protocol. This protocol is claimed to provide strong security and robust privacy protection, while at the same time the usage of resources on tags is optimized. Nevertheless, in this paper we show how the protocol does not achieve any of its intended security objectives; the main result is that the most valuable information stored on the tag, that is, the static identifier ID, is easily recovered even by a completely passive attacker in a number of ways. More precisely, we start by presenting a traceability attack on the protocol that allows tags to be traced. This essentially exploits the fact that the protocol messages leak out at least one bit of the static identifier. We then present a passive attack (named Norwegian attack) that discloses $\lfloor \log_2 L \rfloor$ bits of the ID, after observing roughly O(L) authentication sessions. Although this attack may seem less feasible in retrieving the full 96-bits of the ID due to the large number of eavesdropped sessions involved, it is already powerful enough to serve as a basis for a very effective traceability attack. Finally, our last attack represents a step forward in the use of a recent cryptanalysis technique (called Tango attack [2]), which allows for an extremely efficient full disclosure attack, capable of revealing the value of the whole ID after eavesdropping only a very small number of sessions.