Efficient superdense coding in the presence of non-Markovian noise

TL;DR

This paper demonstrates a superdense coding scheme resilient to noise by exploiting non-Markovian effects, achieving high information transfer efficiency despite significant entanglement loss, with potential applications in secure quantum communication.

Contribution

The authors experimentally show that non-Markovian features can be exploited to maintain high superdense coding efficiency even with highly dephased entangled states.

Findings

Achieved mutual information close to 1.52 bits with 3-state encoding.

Maintained high interference visibility of 99.6%, enabling efficient information transfer.

Demonstrated robustness of superdense coding against non-Markovian noise.

Abstract

Many quantum information tasks rely on entanglement, which is used as a resource, for example, to enable efficient and secure communication. Typically, noise, accompanied by loss of entanglement, reduces the efficiency of quantum protocols. We develop and demonstrate experimentally a superdense coding scheme with noise, where the decrease of entanglement in Alice's encoding state does not reduce the efficiency of the information transmission. Having almost fully dephased classical two-photon polarization state at the time of encoding with concurrence $0.163\pm0.007$, we reach values of mutual information close to $1.52\pm 0.02$ ($1.89\pm 0.05$) with 3-state (4-state) encoding. This high efficiency relies both on non-Markovian features, that Bob exploits just before his Bell-state measurement, and on very high visibility ($99.6\%\pm0.1\%$) of the Hong-Ou-Mandel interference within the experimental set-up. Our proof-of-principle results with measurements on mutual information pave the way for exploiting non-Markovianity to improve the efficiency and security of quantum information processing tasks.