Rate limits in quantum networks with lossy repeaters

TL;DR

This paper investigates how loss in quantum repeater stations impacts the maximum achievable communication rates in quantum networks, providing realistic bounds that account for device imperfections and strategy-dependent effects.

Contribution

It extends existing theoretical bounds to include repeater station loss, modeling its effect on quantum communication rates and strategies in practical network scenarios.

Findings

Maximum rate limited by single station loss regardless of repeater number

Adapted bounds applicable to realistic, lossy quantum repeater networks

Effective channel depends on repeater strategy, influencing achievable rates

Abstract

The derivation of ultimate limits to communication over certain quantum repeater networks have provided extremely valuable benchmarks for assessing near-term quantum communication protocols. However, these bounds are usually derived in the limit of ideal devices and leave questions about the performance of practical implementations unanswered. To address this challenge, we quantify how the presence of loss in repeater stations affect the maximum attainable rates for quantum communication over linear repeater chains and more complex quantum networks. Extending the framework of node splitting, we model the loss introduced at the repeater stations and then prove the corresponding limits. In the linear chain scenario we show that, by increasing the number of repeater stations, the maximum rate cannot overcome a quantity which solely depends on the loss of a single station. We introduce a way of adapting the standard machinery for obtaining bounds to this realistic scenario. The difference is that whilst ultimate limits for any strategy can be derived given a fixed channel, when the repeaters introduce additional decoherence, then the effective overall channel is itself a function of the chosen repeater strategy (e.g., one-way versus two-way classical communication). Classes of repeater strategies can be analysed using additional modelling and the subsequent bounds can be interpreted as the optimal rate within that class.