Unlike purely classical communication, unconditionally secure key distribution is possible if Alice and Bob are both equipped with quantum hardware. The degree to which a protocol needs to be quantum is not only an interesting theoretical question, but also important for practical implementations. Indeed, one may wish to construct cheaper devices, or compensate for device malfunction. In this sense, studying limited resource QKD protocols is an important problem.
One direction to studying this is the semi-quantum model introduced by Boyer et al. in 2007 (PRL 99 140501). Several provably secure semi-quantum protocols were put forth. However, most of these protocols were proven secure in the perfect qubit scenario and not necessarily against practical attacks. Only recently, starting with seminal work of Boyer, Katz, Liss, and Mor in (PRA 96 062335) has research in the field of semi-quantum cryptography considered practical devices and imperfections, such as multi photon sources and imperfect detectors. In this work, we present a new SQKD protocol based on an Extended B92 protocol which is able to counter certain practical attacks. Furthermore, the techniques we use may see broad application to other limited-resource (S)QKD protocols.
Unconditionally secure key distribution is impossible using classical communication only. However, by providing Alice and Bob with quantum capable hardware the task becomes possible. How quantum does a protocol need to be, though, in order to gain this advantage? In 2007, Boyer et al., proposed "semi-quantum key distribution" where only Alice need be quantum while Bob need only limited classical" capabilities. Several protocols were proposed and proven secure in the perfect qubit scenario" but not necessarily against realistic attacks (with one exception being recently published in (PRA 96 062335)). In this paper, we devise a new SQKD protocol and analyze its security against certain practical attacks.