Search

State-of-the-art laser frequency stabilization is limited by miniscule length changes caused by thermal noise. In this work, a cavity-length-insensitive frequency stabilization scheme is implemented using strong dispersion in a $21\,\mathrm{mm}$ long cavity with a europium-ion-doped spacer of yttrium orthosilicate. A number of limiting factors for slow light laser stabilization are evaluated, including the inhomogeneous and homogeneous linewidth of the ions, the deterioration of spectral windows, and the linewidth of the cavity modes. Using strong dispersion, the cavity modes were narrowed by a factor $1.6\cdot 10^5$, leading to a cavity linewidth of $3.0\,\mathrm{kHz}$ and a $Q$ factor of $1.7\cdot 10^{11}$. Frequency stabilization was demonstrated using a cavity mode in a spectral transparency region near the center of the inhomogeneous profile, showing an overlapping Allan deviation below $6\cdot 10^{-14}$ and a linear drift rate of $3.66\,\mathrm{Hz}/\mathrm{s}$. Considering improvements that could be implemented, this makes the europium-based slow light laser frequency reference a promising candidate for ultra-precise tabletop frequency stabilization., Comment: 15 pages, 7 figures

We present a method for blind acoustic parameter estimation from single-channel reverberant speech. The method is structured into three stages. In the first stage, a variational auto-encoder is trained to extract latent representations of acoustic impulse responses represented as mel-spectrograms. In the second stage, a separate speech encoder is trained to estimate low-dimensional representations from short segments of reverberant speech. Finally, the pre-trained speech encoder is combined with a small regression model and evaluated on two parameter regression tasks. Experimentally, the proposed method is shown to outperform a fully end-to-end trained baseline model., Comment: Accepted for publication at IWAENC 2024

Distortion free negative group velocity pulse propagation is demonstrated in a rare-earth-ion-doped-crystal (RE) through the creation of a carefully designed spectral absorption structure in the inhomogeneous profile of Eu:YSO and subsequently inverting it. The properties of the RE system make it particularly well suited for this since it supports the creation of very sharp, arbitrarily tailored spectral features, which can be coherently inverted by a single pulse thanks to the long coherence time of the transition. All together these properties allow for a large time advancement of pulses without causing distortion. A pulse advancement of 10.9% with respect to the pulse full-width-half-maximum was achieved corresponding to a time-bandwidth product of 0.05. This to our knowledge is the largest time-bandwidth product achieved, with negligible shape distortion and attenuation. Our results show that the rare-earth platform is a powerful test bed for superluminal propagation in particular and dispersion profile programming in general, Comment: Main paper 4 pages, 2 figures, total 8 pages 6 figures

Room geometry inference algorithms rely on the localization of acoustic reflectors to identify boundary surfaces of an enclosure. Rooms with highly absorptive walls or walls at large distances from the measurement setup pose challenges for such algorithms. As it is not always possible to localize all walls, we present a data-driven method to jointly detect and localize acoustic reflectors that correspond to nearby and/or reflective walls. A multi-branch convolutional recurrent neural network is employed for this purpose. The network's input consists of a time-domain acoustic beamforming map, obtained via Radon transform from multi-channel room impulse responses. A modified loss function is proposed that forces the network to pay more attention to walls that can be estimated with a small error. Simulation results show that the proposed method can detect nearby and/or reflective walls and improve the localization performance for the detected walls., Comment: 4+1(bib) Pages. Accepted to ICASSP Satellite Workshop - HSCMA 2024

We present a quantum repeater protocol for distributing entanglement over long distances, where each repeater node contains several qubits that can couple to one single-photon emitter. Photons from the emitters perform heralded entanglement generation between qubits in neighboring nodes. The protocol leaves the emitters disentangled from the qubits and photons, thus allowing them to be reused to entangle other qubits. The protocol can therefore be time multiplexed, which increases the rate of generated EPR pairs. Deterministic entanglement swapping and heralded entanglement purification are used to extend the distance of the entanglement and reduce the error of the entangled qubits, respectively. We perform a complete protocol analysis by considering all relevant error sources, such as initialization, two-qubit gate, and qubit measurement errors, as well as the exponential decoherence of the qubits with time. The latter is particularly important since we analyze the protocol performance for a broad range of experimental parameters and obtain secret key rates ranging from $1 \rightarrow 1000$ Hz at a distance of $1000$ km. Our results suggest that it is important to reach a qubit memory coherence time of around one second, and two-qubit gate and measurement errors in the order of $10^{-3}$ to obtain reasonable secret key rates over distances longer than achievable with direct transmission.

Knowing the room geometry may be very beneficial for many audio applications, including sound reproduction, acoustic scene analysis, and sound source localization. Room geometry inference (RGI) deals with the problem of reflector localization (RL) based on a set of room impulse responses (RIRs). Motivated by the increasing popularity of commercially available soundbars, this article presents a data-driven 3D RGI method using RIRs measured from a linear loudspeaker array to a single microphone. A convolutional recurrent neural network (CRNN) is trained using simulated RIRs in a supervised fashion for RL. The Radon transform, which is equivalent to delay-and-sum beamforming, is applied to multi-channel RIRs, and the resulting time-domain acoustic beamforming map is fed into the CRNN. The room geometry is inferred from the microphone position and the reflector locations estimated by the network. The results obtained using measured RIRs show that the proposed data-driven approach generalizes well to unseen RIRs and achieves an accuracy level comparable to a baseline model-driven RGI method that involves intermediate semi-supervised steps, thereby offering a unified and fully automated RGI framework., Comment: Accepted for publication in Forum Acusticum 2023

Rare-earth (RE) ion doped nano-materials are promising candidates for a range of quantum technology applications. Among RE ions, the so-called Kramers' ions possess spin transitions in the GHz range at low magnetic fields, which allows for high-bandwidth multimode quantum storage, fast qubit operations as well as interfacing with superconducting circuits. They also present relevant optical transitions in the infrared. In particular, Er$^{3+}$ has an optical transition in the telecom band, while Nd$^{3+}$ presents a high-emission-rate transition close to 890 nm. In this paper, we measure spectroscopic properties that are of relevance to using these materials in quantum technology applications. We find the inhomogeneous linewidth to be 10.7 GHz for Er$^{3+}$ and 8.2 GHz for Nd$^{3+}$, and the excited state lifetime T$_1$ to be 13.68 ms for Er$^{3+}$ and 540 $\mu$s for Nd$^{3+}$. We study the dependence of homogeneous linewidth on temperature for both samples, with the narrowest linewidth being 379 kHz (T$_2$ = 839 ns) for Er$^{3+}$ measured at 3 K, and 62 kHz (T$_2$ = 5.14 $\mu$s) for Nd$^{3+}$ measured at 1.6 K. Further, we investigate time-dependent homogeneous linewidth broadening due to spectral diffusion and the dependence of homogeneous linewidth on magnetic field, in order to get additional clarity of mechanisms that can influence the coherence time. In light of our results, we discuss two applications: single qubit-state readout and a Fourier-limited single photon source., Comment: 9 pages, 5 figures

A study is presented in which a contrastive learning approach is used to extract low-dimensional representations of the acoustic environment from single-channel, reverberant speech signals. Convolution of room impulse responses (RIRs) with anechoic source signals is leveraged as a data augmentation technique that offers considerable flexibility in the design of the upstream task. We evaluate the embeddings across three different downstream tasks, which include the regression of acoustic parameters reverberation time RT60 and clarity index C50, and the classification into small and large rooms. We demonstrate that the learned representations generalize well to unseen data and perform similarly to a fully-supervised baseline., Comment: Accepted for ICASSP 2023, Camera-ready version

Quantum memories for quantum communication need to be able to store photons for an extended time and then to release them on demand. This can be achieved in atomic frequency comb ensemble based quantum memories by control pulses that transfer the excitation to and from long-lived spin states. However, such pulses can give rise to coherent and incoherent noise due to their interaction with the memory ensemble. In this article, we experimentally demonstrate the ability to switch off the coherent noise from such control pulses during the echo emission in a spin-wave quantum memory, using the linear Stark effect in rare-earth-ion doped crystals. By applying an electric field pulse, the echo emission was coherently switched off prior to the first spin transfer pulse, and the stored data pulse was restored only when both an optical recall pulse and a re-phasing electrical pulse were applied, giving a high degree of control of both desired and undesired emissions. We estimate the effectiveness of this technique by turning off the free induction decay of a narrow ensemble of ions. This technique can thus improve the noise performance of spin-wave storage at the single photon level by quenching coherent optical radiation created by the strong control pulses. The method demonstrated here represents a proof-of-principle that the spin-wave storage scheme can be combined with Stark control. The combined scheme serves as an addition to the toolbox of techniques that can be used to realize a full version of a quantum repeater., Comment: 8 pagrs, 5 figures

A dataset of anechoic recordings of various sound sources encountered in domestic environments is presented. The dataset is intended to be a resource of non-stationary, environmental noise signals that, when convolved with acoustic impulse responses, can be used to simulate complex acoustic scenes. Additionally, a Python library is provided to generate random mixtures of the recordings in the dataset, which can be used as non-stationary interference signals., Comment: Accepted for publication at IWAENC 2022

The estimation of reverberation time from real-world signals plays a central role in a wide range of applications. In many scenarios, acoustic conditions change over time which in turn requires the estimate to be updated continuously. Previously proposed methods involving deep neural networks were mostly designed and tested under the assumption of static acoustic conditions. In this work, we show that these approaches can perform poorly in dynamically evolving acoustic environments. Motivated by a recent trend towards data-centric approaches in machine learning, we propose a novel way of generating training data and demonstrate, using an existing deep neural network architecture, the considerable improvement in the ability to follow temporal changes in reverberation time., Comment: accepted for publication in ICASSP 2022

We use density matrix simulations to study the performance of three distance three quantum error correcting codes in the context of the rare-earth-ion-doped crystal (RE) platform for quantum computing. We analyze pseudothresholds for these codes when parallel operations are not available, and examine the behavior both with and without resting errors. In RE systems, resting errors can be mitigated by extending the system's ground state coherence time. For the codes we study, we find that if the ground state coherence time is roughly 100 times larger than the excited state coherence time, resting errors become small enough to be negligible compared to other error sources. This leads us to the conclusion that beneficial QEC could be achieved in the RE system with the expected gate fidelities available in the NISQ regime. However, for codes using more qubits and operations, a factor of more than 100 would be required. Furthermore, we investigate how often QEC should be performed in a circuit. We find that for early experiments in RE systems, the minimal $[\![5,1,3]\!]$ would be most suitable as it has a high threshold error and uses few qubits. However, when more qubits are available the $[\![9,1,3]\!]$ surface code might be a better option due to its higher circuit performance. Our findings are important for steering experiments to an efficient path for realizing beneficial quantum error correcting codes in early RE systems where resources are limited.

We present two protocols for constructing quantum processor nodes in randomly doped rare-earth-ion crystals and analyze their properties. By varying the doping concentration and the accessible laser tunability, the processor nodes can contain anywhere from only a few tens to almost $1000$ qubits. Furthermore, the average number of qubits each qubit can interact with, denoted by the connectivity, can be partly tailored to lie between just a few and roughly one hundred. We also study how a limited tunability of the laser affects the results, and conclude that a tuning range of $100$ GHz limits the results to roughly $100$ qubits with around $50$ connections per qubit on average. In order to construct an even larger processor, the vision is that several of these quantum processor nodes should be connected to each other in a multi-node architecture via, e.g., optical interfaces or flying qubits in the form of light. Our results are encouraging for establishing the rare-earth-ion-based systems as a quantum computing platform with strong potential and can serve to focus the efforts within the field.

The effect of instantaneous spectral diffusion (ISD) on gate operations in rare-earth-ion-doped crystals is an important question to answer for the future of rare-earth quantum computing. Here we present a microscopic modeling that highlights the stochastic nature of the phenomenon, and use it to investigate ISD errors on single-qubit gate operations. Furthermore, we present a method to estimate the total error from many different error sources by only studying subsystems containing one error source at a time. This allows us to estimate the total ISD error from all non-qubit dopants in the vicinity of a qubit. We conclude that optical pumping techniques must be used to empty the frequency regions around the qubit transitions from absorption (transmission windows) in order to suppress the ISD errors. Despite using such windows, there remains a roughly $0.3\%$ risk that a qubit has an ISD error larger than the error from other sources. In those cases, the qubit can be discarded and its frequency channel can be reused by another qubit. However, in most cases the ISD errors are significantly smaller than other errors, thus opening up the possibility to perform noisy intermediate-scale quantum (NISQ) algorithms despite ISD being present., Comment: Restructured the article by moving content to appendices

Quantum computers based on rare-earth-ion-doped crystals show promising properties in terms of scalability and connectivity if single ions can be used as qubits. Through simulations, we investigate gate operations on such qubits and discuss how gate and system parameters affect gate errors, the required frequency bandwidth per qubit, and the risk of instantaneous spectral diffusion (ISD) occurring. Furthermore, we examine how uncertainties in the system parameters affect the gate errors, and how precisely the system needs to be known. We find gate errors for arbitrary single-qubit gates of $2.1\cdot 10^{-4}$ when ISD is not considered and $3.4\cdot 10^{-4}$ when we take heed to minimize it. Additionally, we construct two-qubit gates with errors ranging from $5\cdot 10^{-4} \rightarrow 3\cdot 10^{-3}$ over a broad range of dipole-dipole interaction strengths.

Length changes due to thermo-mechanical noise originating from, for example, Brownian motion are a key limiting factor of present day state-of-the-art laser frequency stabilization using Fabry-P\'erot cavities. We present a laser-frequency stabilization concept using an optical cavity with a strong slow-light effect to reduce the impact of cavity length changes on the frequency stability. The resulting noise-reduction factor is proportional to the ratio between the light phase and group velocities in the highly dispersive cavity spacer. We experimentally demonstrate a proof-of-principle implementation of this laser-frequency stabilization technique using a rare-earth doped crystalline cavity spacer in conjunction with semi-permanent spectral tailoring to achieve precise control of the dispersive properties of the cavity. Compared to the same setup in the absence of the slow-light effect a reduction in frequency sensitivity of four orders of magnitude was achieved., Comment: Published version, 24 pages, 7 figures

Several platforms are being considered as hardware for quantum technologies. For quantum computing (QC), superconducting qubits and artificially trapped ions are among the leading platforms, but many others also show promise, e.g. photons, cold atoms, defect centers including Rare-Earth (RE) ions. So far, results are limited to the regime of noisy intermediate scale qubits (NISQ), with a small number of qubits and a limited connectivity, and it is likely that future QC hardware will utilize several existing platforms in different ways. Thus, it currently makes sense to invest resources broadly and explore the full range of promising routes to quantum technology. Rare-earth ions in solids constitute one of the most versatile platforms for future quantum technology. One advantage is good coherence properties even when confined in strong natural traps inside a solid-state matrix. This confinement allows very high qubit densities and correspondingly strong ion-ion couplings. In addition, although their fluorescence is generally weak, cavity integration can enhance the emission greatly and enable very good connections to photonic circuits, including at the telecom wavelengths, making them promising systems for long-term scalability. The primary aim of this roadmap is to provide a complete picture of what components a RE quantum computer would consist of, to describe the details of all parts required to achieve a scalable system, and to discuss the most promising paths to reach it. In brief, we find that clusters of 50-100 single RE ions can act as high fidelity qubits in small processors, occupying only about (10 nm)^3. Due to the high capacity for integration of the RE systems, they be optically read out and connected to other such clusters for larger scalability. We make suggestions for future improvements, which could allow the REQC platform to be a leading one., Comment: 47 pages, 12 figures. Adapted from a project report

Accurate and efficient quantum control in the presence of constraints and decoherence is a requirement and a challenge in quantum information processing. Shortcuts to adiabaticity, originally proposed to speed up slow adiabatic process, have nowadays become versatile toolboxes for preparing states or controlling the quantum dynamics. Unique shortcut designs are required for each quantum system with intrinsic physical constraints, imperfections, and noises. Here, we implement fast and robust control for the state preparation and state engineering in a rare-earth ions system. Specifically, the interacting pulses are inversely engineered and further optimized with respect to inhomogeneities of the ensemble and the unwanted interaction with other qubits. We demonstrate that our protocols surpass the conventional adiabatic schemes, by reducing the decoherence from the excited state decay and inhomogeneous broadening. The results presented here are applicable to other noisy intermediate scale quantum systems.

We present an extension of the atomic frequency comb protocol that utilizes the Stark effect to perform noise-free, on-demand, control. An experimental realization of this protocol was implemented in the Pr$^{3+}$:Y$_2$SiO$_5$ solid-state system, and a recall efficiency of 38\% for a 0.8 $\mu$s storage time was achieved. Experiments were performed with both bright pulses as well as weak-coherent states, the latter achieving a signal-to-noise ratio of $570 \pm 120$ using input pulses with an average photon number of $\sim 0.1$. The principal limitation for a longer storage time was found to be the minimum peak width attainable for Pr$^{3+}$:Y$_2$SiO$_5$. We employ an adaptation of an established atomic-frequency comb model to investigate an on-demand, wide-bandwidth, memory based on Eu$^{3+}$:Y$_2$SiO$_5$. From this we determine that a storage time as long as 100 $\mu$s may be practical even without recourse to spin-wave storage., Comment: Published version. 9 pages, 7 figures

High-fidelity qubit initialization is of significance for efficient error correction in fault tolerant quantum algorithms. Combining two best worlds, speed and robustness, to achieve high-fidelity state preparation and manipulation is challenging in quantum systems, where qubits are closely spaced in frequency. Motivated by the concept of shortcut to adiabaticity, we theoretically propose the shortcut pulses via inverse engineering and further optimize the pulses with respect to systematic errors in frequency detuning and Rabi frequency. Such protocol, relevant to frequency selectivity, is applied to rare-earth ions qubit system, where the excitation of frequency-neighboring qubits should be prevented as well. Furthermore, comparison with adiabatic complex hyperbolic secant pulses shows that these dedicated initialization pulses can reduce the time that ions spend in the excited state by a factor of 6, which is important in coherence time limited systems to approach an error rate manageable by quantum error correction. The approach may also be applicable to superconducting qubits, and any other systems where qubits are addressed in frequency., Comment: 15 pages, 9 figures