Theory of entanglement and entanglement-assisted communication

Janis Nötzel, Stephen DiAdamo

Shuhong Hao, Haowei Shi, Wei Li +3 more

Entanglement underpins a variety of quantum-enhanced communication, sensing, and computing capabilities. Entanglement-assisted communication (EACOMM) leverages entanglement preshared by communicating parties to boost the rate of classical information transmission. Pioneering theory works showed that EACOMM can enable a communication rate well beyond the ultimate classical capacity of optical communications, but an experimental demonstration of any EACOMM advantage remains elusive. In this Letter we report the implementation of EACOMM surpassing the classical capacity over lossy and noisy bosonic channels. We construct a high-efficiency entanglement source and a phase-conjugate quantum receiver to reap the benefit of preshared entanglement, despite entanglement being broken by channel loss and noise. We show that EACOMM beats the Holevo-Schumacher-Westmoreland capacity of classical communication by up to 16.3%, when both protocols are subject to the same power constraint at the transmitter. As a practical performance benchmark, we implement a classical communication protocol with the identical characteristics for the encoded signal, showing that EACOMM can reduce the bit-error rate by up to 69% over the same bosonic channel. Our work opens a route to provable quantum advantages in a wide range of quantum information processing tasks.

Satyaki Manna, Ankush Pandit, Anubhav Chaturvedi +1 more

We investigate the quantum advantage that can arise in typical two-party communication scenarios, where the sender and the receiver are allowed to share prior correlations. Focusing on communication tasks constrained by the distinguishability of the sender's inputs, we demonstrate that entanglement-assisted communication with both classical and quantum message can outperform classical communication supplemented with shared randomness. We begin by developing a general framework for communication tasks with pre-shared correlations. Within this framework, we establish an equivalence among entanglement-assisted classical communication, entanglement-assisted quantum communication, and quantum communication, showing that no hierarchy exists between these three paradigms. We then investigate the scenario where the receiver has no input and prove that no advantage can arise in this case. However, an advantage in the entanglement-assisted setting emerges once additional constraints are imposed on the dimension of the communicated message. This further highlights the superiority of entanglement-assisted classical communication over standard quantum communication. Then we demonstrate several tasks where the entanglement-assisted protocol using one-bit communication proves to be advantageous over classical communication. Finally, by constructing an explicit class of communication tasks, we show that a non-maximally entangled states outperform the maximally entangled state as a pre-shared resource between the communicating parties.

Mark M. Wilde

Ivan B. Djordjević

Entanglement assisted communication is advocated by numerous authors as an alternative to classical communication offering significant improvement in channel capacity, in particular in noisy regime. In all those papers it is always assumed that entanglement can be distributed without any imperfections, except for attenuation. We demonstrate that under imperfect pre-shared entanglement distribution, assuming that entanglement distribution channel is modeled as a noisy and lossy Bosonic channel, the entanglement assisted communication can be inferior compared to the classical communication, depending on the parameters of the distribution channel. We identify the conditions under which entanglement assistance can still provide an advantage over the classical case. In particular, when both communication and entanglement distribution channels are not used for entanglement assisted communication but rather for classical transmission instead then the classical capacity is always higher than the entanglement assisted capacity. We also study the entanglement assisted communication under the strong atmospheric turbulence effects.

Mark M. Wilde