Detector-device-independent QKD: security analysis and fast implementation

TL;DR

This paper analyzes the security of detector-device-independent QKD, clarifies its assumptions, and demonstrates a fast experimental implementation capable of secure key exchange over long distances.

Contribution

It provides the first security analysis of DDI-QKD and presents a high-speed experimental demonstration over 90 km.

Findings

DDI-QKD's security can be established under reasonable assumptions.

A 625 MHz implementation successfully exchanges keys over 90 km.

DDI-QKD is not equivalent to MDI-QKD but offers practical advantages.

Abstract

One of the most pressing issues in quantum key distribution (QKD) is the problem of detector side- channel attacks. To overcome this problem, researchers proposed an elegant "time-reversal" QKD protocol called measurement-device-independent QKD (MDI-QKD), which is based on time-reversed entanglement swapping. However, MDI-QKD is more challenging to implement than standard point- to-point QKD. Recently, an intermediary QKD protocol called detector-device-independent QKD (DDI-QKD) has been proposed to overcome the drawbacks of MDI-QKD, with the hope that it would eventually lead to a more efficient detector side-channel-free QKD system. Here, we analyze the security of DDI-QKD and elucidate its security assumptions. We find that DDI-QKD is not equivalent to MDI-QKD, but its security can be demonstrated with reasonable assumptions. On the more practical side, we consider the feasibility of DDI-QKD and present a fast experimental demonstration (clocked at 625 MHz), capable of secret key exchange up to more than 90 km.