A digital twin of atomic ensemble quantum memories

Elizabeth Jane Robertson †, Benjamin Maaß †, Konrad Tschernig, Janik Wolters · 2025

Read A digital twin of atomic ensemble quantum memories by Elizabeth Jane Robertson †, Benjamin Maaß †, Konrad Tschernig, Janik Wolters (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-20403.txt.

Introduction

Quantum networks will need memory devices that hold light based information without losing it, and testing real devices is slow and costly. Elizabeth Robertson and three co-authors build a digital twin of atomic ensemble memories in 2025, a software model that copies how the real hardware behaves, losses and noise included. They give model forms for several leading memory designs.

The twin also reports the scores that matter most for judging a memory's worth. To show it in action, the team ran a memory assisted quantum token scheme inside the model. Because the twin matches the language of network simulators, it can plug into wider studies. The authors offer it as a cheap test bed for future quantum networks.

Catalog record

CREATORElizabeth Jane Robertson †, Benjamin Maaß †, Konrad Tschernig, Janik Wolters
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE95 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-20403

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A digital twin of atomic ensemble quantum memories Elizabeth Jane Robertson †, Benjamin Maaß †, Konrad Tschernig, Janik Wolters Abstract Accurate performance estimation of experimentally demonstrated quantum memories is key to understand the nuances in their deployment in photonic quantum networks. While several software packages allow for accessible quantum simulation, they often do not account for the loss and noise in physical devices. We present a framework for modeling ensemble-based atomic quantum memories using the quantum channel formalism. We provide a Kraus matrix representation of several experimentally implemented state-of-the art quantum memories and give an overview of their most important performance metrics. To showcase the applicability of this approach, we implement a memory-assisted quantum token protocol within our simulation framework. Our digital twin model is readily extensible to other memory implementations and easily compatible with existing frameworks for performance simulation of experimental quantum networks. Keywords — quantum memory, quantum network, digital twin, performance simulation † † footnotetext: † The authors contributed equally to this work. † † footnotetext: * Corresponding author, Email: [email protected] 1 Introduction The vision of a quantum internet [ 1 ] has driven quantum technology research over the last decades. The backbone of such a network is entanglement distribution between remote network nodes that are connected via photonic quantum channels [ 2 ] . However, optical losses in fibers severely limit the possible scale of entanglement distribution and create the need for quantum repeaters. In this context, quantum memories are critical components of any repeater-based transfer of quantum information. Their application in buffering, synchronizing and conditioning of photonic quantum information requires an in-depth knowledge of the inner workings of quantum memories [ 3 , 4 ] . Equally as important is the ability to understand the performance of different quantum memory platforms, experimental implementations and compatibilities with other photonic resources, all within the context of multi-layered network architectures. Theoretical investigations of ensemble-based quantum memories on the one hand are well-established [ 5 , 6 ] but often remain very specialized in their scope. On the other hand, recent years have shown remarkable progress in the development of tool-kits for photonic quantum simulation. Prominently, Strawberry fields [ 7 ] , Perceval [ 8 ] and Piquasso [ 9 ] have emerged as tools for Gaussian- and discrete Boson sampling and photonic quantum computing. In addition, tools for simulating open quantum systems QuTiP [ 10 ] and gate-based computing Qiskit [ 11 ] with extension module for optics SOQCS [ 12 ] have been published. However, there is a lack of open source, close-to-experiment plug-and-play software packages to simulate real quantum communication devices taking into account noise, loss and decoherence processes which requires significant manual customization in presently available simulation frameworks. In this work we develop a general model of atomic ensemble-based quantum memories, taking loss-, noise- and decoherence processes into account, and demonstrate the plug-and-play software implementation in several test-scenarios. First, we describe the basic working principle of ensemble-based quantum memories and introduce their most important performance characteristics 2 . Second, after introducing the description of a quantum memory within the channel formalism 3 we demonstrate the functionality of the simulation and apply it in a Mach-Zehnder configuration. Furthermore, we introduce simulation building blocks as digital twins of state-of-the-art vapor memories which allows for straightforward performance comparisons. Finally, we apply the digitally twined memories to a quantum token protocol and benchmark the memory performances against the security of the protocol 5 . 2 Quantum Memories Fundamentally, a memory is a device that stores some type of information encoded in a propagating or short-lived state. The information is transferred to a storage medium by encoding it in a long-lived storage state and retrieved at a later time with a certain efficiency and fidelity. The term efficiency refers to the probability of a successful memory operation versus the number of trials. The fidelity of a memory is defined as the similarity between the input and the retrieved information. In the case of quantum states to be stored the re-encoding of information in a storage medium is fundamentally difficult as it entails a projective measurement of the quantum state which changes the properties of the state itself. Specifically, an unknown quantum state cannot be reconstructed from a single measurement. Crucially, this means that storing quantum information requires storing the entire state and not only its projections onto some measurement basis during the storage process. The challenge in constructing such a system lies in the fundamental impossibility to create identical copies of arbitrary quantum states [ 13 ] . Figure 1 : A typical photon arrival time histogram for an optically controlled memory experiment. The upper panels show the envelopes of the signal and the control field fed into the memory. The overlap of the control pulse with the signal pulse within the medium at time 0.22 maps the signal pulse to a collective spin-(orbital) wave. The experimental result (bottom panel) shows some leakage at this point in time from unstored light. After the storage time \tau_{\mathrm{s}} the control pulse is re-applied and the signal field is retrieved. In this work we focus on photonic quantum memories in particular, i.e. memory devices that store information encoded in states of light. The most straight-forward way of storing light is in a fiber loop or Herriot cell. An incoming state of light is guided onto a long or, in the case of a fiber loop, closed light path that delays the transmission of the state. These implementations usually provide high fidelity and broadband operation but come with fixed read-in and read-out intervals [ 14 ] . This work concentrates on optically controlled ensemble-based memories. In these implementations a light field is used to exert control over the interaction between the light field to be stored and some atomic ensemble that serves as storage medium. During the read-in process the photonic state is mapped onto a coherent light-matter state, sometimes referred to as spinwave or (dark state) polariton under an ideally unitary (time-reversible) transformation. This allows one to reverse the read-in process after the storage time \tau_{\mathrm{s}} and retrieve the initial photonic state. During the storage time the state undergoes some decoherence processes that heavily depend on the specific implementation. However, since both the read-in and read-out processes are (ideally) unitary the quantum coherence between the initial state and the retrieved state is conserved. An example of the storage and retrieval of a coherent state from an ensemble vapor memory can be seen in Figure 1 . The signal pulse (red) is mapped onto the spinwave by applying the first control pulse (orange). After time \tau_{\mathrm{s}} , a second control pulse is sent into the atomic ensemble, and the state is retrieved from the memory (light blue). There is a wide variety of different memory implementations with different properties that determine their suitability for application. In the next section, we describe the most important key characteristics on the basis of which ensemble quantum memories are compared. 2.1 Characteristics and figures of merit 2.1.1 Efficiency The efficiency of a quantum memory describes the probability of retrieving a photon from the memory after a certain time, i.e. generally the memory efficiency is a function of the time

Read the complete text