We study Bayesian decoding of finite-energy Gottesman–Kitaev–Preskill (GKP) qubits from multi-shot homo dyne records under coherent displacement channels, in which all shots of a trial share one latent shift. Our first contribution is an α-marginalised Bayes-optimal decoder, validated shot-for shot against the exact likelihood computed from the simulator's own wavefunction/Fock pipeline: the textbook IID Gaussian mixture maximum-likelihood decoder is 3.2–3.6× worse across σch = 0.4–0.7 (20 seeds), and the failure persists down to shot-shift correlations of ρ ≈ 0.5. Our second contribution is QIFE (Quantum Information Field Explorer), a closed-form Wiener-filter decoder on fixed 36-dimensional Fourier-modular shot statistics that uses no likelihood at all: it trains from a labelled calibration set in 0.67ms (130–240× faster than matched-budget neural decoders, which it also outperforms at every operating point tested) and decodes in 1.3µs per record (∼ 750× faster than the α-marginalised decoder). Third, an equal-calibration-budget comparison: when the same 300 labelled records used to train QIFE are instead used to estimate the channel parameters by grid maximum likelihood, the resulting plug-in Bayes decoder outperforms QIFE at every channel-drift cell we test. We therefore do not recommend noise-model-free decoding when even a coarse loss-aware likelihood is available; its niche is likelihood-free settings and latency-critical post processing. All results are reproducible from an open-source package with an exact-likelihood validation suite.
This work presents a novel micromachined, seamless sapphire–silicon hybrid evanescent-mode cavity architecture incorporating selectively engineered coupling channels and flux-tunable external coupling via a SQUID. The proposed design enables two simulated operating states within a single device: an ultra–high Q memory configuration and a high-speed readout configuration with intrinsic Purcell filtering, which is expected to enhance readout fidelity while preserving the coherence of the stored quantum information. Unlike prior 3D and multi-mode cavity approaches, the design employs a distributed embedded-capacitance network that confines electric fields into ultra–low-loss dielectric regions while distributing magnetic fields to an air-filled enclosure, thus substantially reducing simulated surface and dielectric losses beyond reported architectures. Moreover, the structure employs contactless superconducting metallization between two wafers, eliminating seam losses typically observed in micromachined resonators. The designed selective field-coupling strategy enables pure magnetic coupling through the lower silicon substrate, enabling fast readout, while pure electric coupling through the upper sapphire interface ensures strong qubit–cavity in teraction, without mixed-coupling artifacts. High-fidelity mixed electromagnetic and circuit-level modeling predicts an intrinsic-Q exceeding 108 at 5554.5 MHz. The flux-tunable asymmetric DC-SQUID enables a dynamic external-Q reconfiguration from 109 (memory mode) to 104 (readout mode). When coupled to a 6981.7 MHz transmon qubit, the architecture provides a nearly six-order-of-magnitude simulated Purcell-limited qubit lifetime improvement and achieves 66% simulated memory efficiency at 1 ms under realistic coherence assumptions. These simulated results show that the proposed seamless, capacitively engineered architecture can work as a promising, high-coherence, and rapidly reconfigurable platform for scalable superconducting quantum systems.
Practical NV-center magnetometry requires both accurate modeling of open quantum dynamics and robust control under environmental drift. However, commonly used perturbative descriptions inadequately capture dissipative processes, while existing control strategies remain individually limited: open-loop protocols cannot adapt to unknown frequency variations, Lyapunov feedback typically converges only to a neighborhood of the optimal sensing state, and reinforcement learning alone lacks stability guarantees and requires extensive training. Here we develop a Lindblad-based framework for one effective NV center that explicitly incorporates relaxation and dephasing; the analytical model does not describe a collective many-NV ensemble. We introduce a dissipation-triggered feedback mechanism that restores high-sensitivity states following spontaneous emission events. Building on this model, we propose a hybrid control architecture that combines Lyapunov stabilization with reinforcement-learning-based adaptive correction to compensate unknown Hamiltonian drift. Robustness tests over detuning and microwave-amplitude variations show that Hybrid attains the highest QFI, substantially lower cross-realization dispersion than Pure RL and Robust GRAPE, faster fixed-budget learning than Pure RL, and substantially lower offline dynamical-propagation cost with smoother control modulation than Robust GRAPE. This physics-informed and data-driven strategy provides a robust and resource-efficient approach for maintaining high quantum Fisher information in realistic NV sensing environments.
We have investigated the entanglement distribution and quantum teleportation in quantum passive optical networks (QPONs), where an entangled state is distributed between the central node and multiple end users through lossy fibers and passive optical splitter/combiner. The entanglement degradation is quantified using logarithmic negativity, and teleportation fidelity both in non-assisted and assisted strategies. We find that non-assisted teleportation fidelity rapidly drops below the classical limit as the number of users increases (possible up to 5 users), while assisted teleportation remains robust against both network size and photon loss provided that the distributed entangled resource has sufficiently high fidelity. These results highlight the fundamental limitations of bipartite schemes and establish cooperative assisted strategies as essential primitives for scalable fiber-to-the-home (FTTH) quantum access networks and the quantum internet with several end users.
The single-photon dual-rail qubit is popular for photonic quantum computing and communications, both due to gates being realizable with linear optics (LO) and photon detection, as well as easy interfaces with matter quantum memories. Two-qubit operations, i.e., two-qubit gates followed by the measurement of one (Type-I fusion) or both (Type-II fusion) qubits, are, however, probabilistic. While well-known LO circuit realizations for a few standard two-qubit operations exist, translating an arbitrary multi-qubit quantum operation to a LO circuit, and vice versa, remain unclear. Given any $k$-qubit stabilizer operation, we provide a prescriptive—albeit potentially suboptimum in terms of probability of success—LO circuit design and an associated ‘success’ pattern of photon-detection outcomes. We introduce new variations of Type-I fusions and their stabilizer-operation descriptions. These provide a rich set of tools for resource-efficient photonic graph-state preparation. We provide LO circuits and stabilizer-operation descriptions for general Type II fusions, which project two qubits onto an arbitrary pair of opposite-phase Bell states. Finally, we provide graph theoretic rules for the success and failure outcomes for both type I and II fusions, as well as a few examples of multi-qubit stabilizer operations. Our results thereby eliminate the need for quantum optics calculations for LO circuit design for realizing general stabilizer operations.
Quantum Key Distribution (QKD) is a cryptographic solution that leverages the properties of quantum mechanics to be resistant and secure even against an attacker with unlimited computational power. Satellite-based links are important in QKD because they can reach distances that the best fiber systems cannot. However, links between satellites in Low Earth Orbit (LEO) and ground stations have a duration of only a few minutes, resulting in the generation of a small amount of secure keys. In this context, we investigate the optimization of the information reconciliation step of the QKD post-processing in order to generate as much secure key as possible. As a first step, we build an accurate model of the downlink signal and Quantum Bit Error Rate (QBER) during a complete satellite pass, which are time-varying due to three effects: (i) the varying link geometry over time, (ii) the scintillation effect, and (iii) the different signal intensities adopted in the Decoy-State protocol. Leveraging the a-priori information on the instantaneous QBER, we improve the efficiency of information reconciliation (i.e., the error correction phase) in the Decoy-State BB84 protocol, resulting in a secure key that is almost 3 no computational or hardware complexity overhead.
This paper proposes an optimization model for the design of quantum communication infrastructures (QCI) with the aim of minimizing capital expenditure (CAPEX), focusing on the costs associated with the deployment of quantum links and trusted repeater nodes (TRNs). The model is formulated using an integer linear programming (ILP) program that incorporates physical and operational constraints to ensure efficient resource allocation. The simulation results obtained using an ILP solver show that increasing wavelength availability leads to significant reductions in CAPEX while consistently producing solutions that satisfy all constraints. However, because ILP becomes computationally expensive for large-scale networks, we propose two genetic algorithms ($GA_{1}$ and $GA_{2}$) designed to improve scalability. Both $GA_{1}$ and $GA_{2}$ offer competitive computation times compared with the ILP solver and provide near-optimal solutions compared with the ILP solution. Overall, this approach is effective for small, medium, and large-scale QCI deployments, offering a scalable and practical solution suitable for real-world implementation. For example, in the small network topology with 9 nodes and 9 requests, we observe a reduction of CAPEX up to 33% when the number of available wavelengths increases from 1 to 2, and we note that genetic algorithms in a large topology, such as NET-4 network, reduce CAPEX by 34.6% ($GA_{1}$) and 21.1% ($GA_{2}$) compared to the benchmarks.
Capsule Network (CapsNet) provides a richer hierarchical representation than conventional neural networks through vector-shaped capsules and inter-capsule routing. However, most CapsNet designs still rely on classic matrix-multiplication-based vote transformations, which limit the formation of the prediction vector to linear mappings in Euclidean space. In this paper, we propose the Hybrid Quantum Capsule Network (HQCapsNet), an architecture that replaces classical vote transformations with a Quantum Transformation Circuit (QTC) based on a Parameterized Quantum Circuit (PQC). We introduce QTC in two schemes: class-wise QTC and fully-connected QTC. Furthermore, the routing agreement mechanism is extended via Hybrid Quantum Routing, leveraging various quantum similarity metrics. The evaluation was conducted on several binary image classification datasets, with MNIST, TMNIST, and MNIST-M used as the primary datasets. The experimental analysis included benchmarking of quantum embedding variations, the PQC ansatz, qubit-count analysis, circuit depth, and routing iterations. The experimental results show that fully-connected QTC is the best configuration, achieving consistent accuracy improvements over class-wise QTC, with the most significant improvement from 92.86% to 97.62% on the MNIST-M dataset. Compared to the classical CapsNet, HQCapsNet also achieves higher performance across all primary datasets, achieving 99.95% accuracy on MNIST and competing with quantum-classical models on comparable split data. These results indicate that integrating quantum computing at the voting transformation step is a promising approach to improving CapsNet's representation and generalization capabilities.
We develop an exact error probability framework for quantum free space optics (FSO) communication over atmospheric channels with turbulence and pointing errors under a physically admissible bounded transmissivity model. The proposed framework is grounded in the fundamental physical observation that, in a quantum optical channel, the random channel quantity is the transmissivity itself, namely, the fraction of transmitted photons collected at the receiver, and must therefore take values only in the physically admissible interval [0,1]. This constraint is generally violated when classical turbulence models are interpreted directly as transmissivity models in quantum settings. Accordingly, we model the turbulence-induced transmissivity as a Beta distributed random variable over (0,1), which has recently been shown to provide accurate fitting under different turbulence conditions while preserving physical admissibility. We then integrate deterministic atmospheric loss and pointing errors into a unified equivalent transmissivity model and derive exact expressions for its probability density function and moment generating function. Building upon this statistical characterization, we derive novel exact average error probability expressions for binary modulation under both quantum and classical receivers. Finally, we verify the analysis using Monte Carlo simulations, quantify the impact of turbulence and pointing errors, and demonstrate the performance advantage of quantum optimal detection in photon-limited quantum free space optics (FSO) channels.
By simulating the selective focus characteristics of human visual attention, the attention network has greatly improved the model's ability to capture key features, promoting significant progress in natural language processing, computer vision and other fields. As a typical attention mechanism, the Squeeze-and-Excitation (SE) module finely adjusts the feature representation of the neural network by re-weighting the features between channels, further enhancing the performance and generalization ability of the model. In this paper, we introduce Quantum Squeeze-and-Excitation (QSE) Networks, a pioneering approach that enhances the excitation module of classical SE networks using quantum computing. Our method simplifies the model's complexity and boosts performance by employing quantum amplitude coding for data encoding, significantly reducing the parameter count of fully connected layers in classical SE module. To optimize the quantum circuits within the QSE module, we explore five different Controlled-NOT (CNOT) gate topologies: circular, linear, star, tree, and mesh. Experimental results show that, after 100 training rounds, the accuracy of our proposed linear topology-based QSE ResNet-18 on the CIFAR-10 dataset reached $85.00\%$ in no noise case, while the classical SE ResNet-18 was only $82.20\%$. Contrary to the common perception of noise as a disruptive and unavoidable challenge, we demonstrate that, through specific topology structures and a hybrid quantum-classical QSE design, quantum noise does not need to be deliberately avoided. Instead, we leverage quantum noise to enhance the robustness of QSE networks and mitigate the Local Optima Trapping problem caused by circuit redundancy by four quantum noise models-amplitude damping, depolarizing, phase-flip, and bit-flip noise. Experimental results show that under four noise models, the QSE accuracy of the five different topological results we proposed can reach a performance of more than $85.00\%$. The inevitable quantum noise inherently serves as an advantage in the proposed QSE.
Solving partial differential equations (PDEs), which is pervasive in science and engineering, is emerging as a promising application area for quantum computing because it can be reduced to Hamiltonian simulation. Unlike quantum chemistry, Pauli-term expansion is not useful for a PDE Hamiltonian, since it ends up with an exponential number of Pauli terms. Recently, it has been shown that, by diagonalizing tensor products of ladder operators, such a Hamiltonian can be mapped into a scalable circuit with a polynomial number of multi-controlled gates. When optimized by an industry-grade circuit compiler, the resulting circuit is realized with a quadratic number of CX gates with respect to the number of qubits. To further optimize PDE circuits, we propose Multilevel Gate Set Optimization (MGSO), an approach for identifying effective decomposition methods for high-level gates. In MGSO, we define multiple levels of gate sets, from higher to lower. At each level, we optimize the circuit and then lower it using decomposition methods carefully selected based on circuit structure to maximize optimization opportunities at subsequent levels. Applying MGSO to optimize the PDE circuits, we achieved a quadratic reduction in the number of CX gates with a linear increase in the number of qubits.
Classical simulation remains a practical foundation for developing and evaluating quantum circuits, and state-vector methods are still among the most important exact simulation techniques. In this setting, gate fusion is a widely used optimization that combines multiple gates into a larger operator to reduce repeated state-vector traversal and improve execution efficiency. However, existing fusion methods are often driven by local heuristics or linear gate order, which can miss larger legal fusion opportunities allowed by circuit dependencies. To address this issue, we propose a DAG-aware gate fusion method that builds legal fused operators directly from the circuit dependency graph through topological expansion, a configurable fusion-size bound, and a non-blocking rule. Using the pyquafu state-vector simulator as the execution backend, the method delivers strong end-to-end performance on representative benchmarks and yields high compression ratios. The results also highlight that fusion quality alone is not sufficient: the realized speedup depends on how well the fused blocks match the execution characteristics of the simulator backend.
The Al/AlOx/Al system is widely used for the fabrication of Josephson junctions, which constitute the central element of superconducting qubits. The process parameters for growing ultra-thin AlOx barriers by thermal oxidation in-between physical vapor deposition (PVD) runs of Al films are well understood, but the resulting barriers present several drawbacks that limit performance such as thickness inhomogeneity and oxygen deficiency that leads to high defect densities in its bulk and interfaces. In this work, we present advances in the development of an alternative fabrication process for Al/Al2O3/Al structures by using thermal atomic-layer deposition (ALD) able to produce high-quality Al2O3 films with thickness close to 2 nm. We focus on reducing the low-quality interfacial oxide between Al and Al2O3 by using well-controlled wet chemical etching to eliminate the surface oxide layer that forms on the bottom PVD-deposited Al film. Additionally, the ALD deposition process for Al2O3 is also optimized by adjusting the length of the precursor pulses and purges to promote the nucleation of the Al2O3 first cycles on the Al surface. The effect of the processing conditions was observed through spectroscopic ellipsometry, electrical measurements at room temperature, and modeling of charge transport mechanisms on Al/Al2O3/Al structures. The electrical characterization revealed nonlinear J–V characteristics consistent with tunneling dominated transport, while the transport modeling enabled estimation of effective barrier thickness, interfacial layer thickness, barrier normal resistance, and critical current. The results show that interface engineering and ALD process optimization can effectively reduce the interfacial oxide contribution and improve barrier uniformity, providing useful insights for the development of Al/Al2O3/Al tunnel junctions for superconducting quantum circuit applications.
This article presents a comprehensive survey of the current frontier in quantum computing for computational sciences, evaluating the technical requirements to translate theoretical asymptotic speedups into practical utility in the areas of chemistry, biochemistry, and materials science. We review foundational algorithms, including the quantum Fourier transform, quantum phase estimation, and the quantum linear-system solver, along with variational heuristics such as variational quantum eigensolver and quantum approximate optimization algorithm. Particular emphasis is placed on near-optimal Hamiltonian simulation frameworks, specifically qubitization and quantum signal processing. Furthermore, we shed light on the current advances in quantum error correction codes, quantum hardware, and quantum software platforms. We then provide a comprehensive review of the application of quantum algorithms in four computational science domains that collectively represent the most compelling near-term targets for quantum advantage. This article aims to provide a clear and balanced perspective on the current state of the field and its future potential for advancing computational science.
Engineering quantum circuits that use minimal resource requirements is essential for suppressing noise-induced errors and enhancing the performance of quantum processors. Here, we propose minimal-complexity hardware constructions of Clifford circuits for implementing new two-qubit Clifford gates, effectively expanding the available Clifford circuit library. The circuits are realized through engineered coherent phonon-mediated interactions between two subsets of SiV$^{-}$ centers, each of which is independently controlled by microwave driving fields that modulate the effective Hamiltonian. In addition, we demonstrate that under appropriate microwave driving parameters, this setup can significantly reduce circuit depth, enabling more efficient gate execution and improved complexity speedup. Moreover, we investigate the impact of pulse-area errors in the proposed circuits to evaluate their robustness against systematic errors. The protocol shows potential compatibility with scalable quantum hardware for implementing advanced algorithms in future quantum technologies.
We propose a novel design of Ge1-xSnx-on-Si single-photon avalanche photodiodes (SPADs) that aim to enhance the fill factor (FF) and minimize noise at room temperature. The device consists of a n(+)/i-well dot structure designed to eliminate the need for guard rings and multidot or array configurations typically used to enhance the active area. This study considers three distinct concentrations of Sn (4%, 6%, and 8%) in the GeSn active layer and investigates their effect on performance metrics. The impact of the threading dislocation density in defective GeSn on the dark count rate (DCR) is also examined. The results show that a high Sn concentration has a notable impact on dark current. However, the DCR, single-photon detection efficiency (SPDE), and noise-equivalent power exhibit less sensitivity to variations in Sn concentration. All devices exhibit exceptionally low dark currents (<75 pA), a high GeSn absorption coefficient, and a high triggering probability (>95%), yielding a significant SPDE (>82.2%). Furthermore, the minimal DCR value (<0.0095 Mcps) combined with a high SPDE results in a low noise equivalent power (<0.02 fWHz(-0.5)) and high detectivity at lambda = 1.55 mu m and V-EX = 5 V. The GeSn-on-Si SPADs here address the challenges faced by current Ge-on-Si, SiGe, InGaAs/AlGaAsSb, and InGaAs/InP SPADs regarding low operating temperatures, showcasing their promise for quantum photonics and communication applications at room temperature (T = 300 K).
Hole spin qubits in silicon nanostructures offer fast, all-electrical control through electric dipole spin resonance, yet their performance strongly depends on device geometry. In this work, optimization criteria of Rabi frequency of single-hole spin qubits in silicon-on-insulator quantum dots are identified by combining electrostatic and k & sdot;p simulations with a perturbative model of the Rabi frequency linear in the magnetic field magnitude. Starting from a reference device deeply studied in the literature, the dominant terms governing the Rabi frequency are recalled, remarking the key role of the component of the RF electric field parallel to the long axis of the nanowire cross section. Guided by such insights, two alternative devices designed to enhance this field component are proposed, leveraging the gate geometries and the crystal orientation. In the strain-free case, a more than fourfold improvement in the peak Rabi frequency over the reference device is achieved, partially offset by a shorter dephasing time, yet still providing an improvement of more than 10% in the quality factor Q(2)(& lowast;).The analyses are then repeated with uniform in-plane biaxial strain. Again, higher Rabi frequenciesare generally obtained, accompanied by shorter dephasing times, resulting in improvements in Q(2)(& lowast;) that can approach one order of magnitude, although strongly dependent on the strain value.
Most existing quantum teleportation schemes do not consider the memory effect of noise, which is becoming increasingly serious in practical quantum communication. In this paper, we propose two quantum teleportation protocols for noisy memory channels: one based on pre-flipping (PF) and another incorporating pre-flipping with environment-assisted measurement (PF-EAM). In the PF protocol, a pre-flipping operation is applied to the entangled qubits before entanglement distribution to enhance robustness against noise. A recovery operator is then used to reverse the pre-flipping operation. In the PF-EAM protocol, environment-assisted measurement (EAM) is added to the entanglement distribution process, further improving teleportation fidelity. Since the pre-flipping operator in the two protocols is a fixed unitary operator, our protocols exclude additional parameters, which is unlike weak measurement (WM) schemes that rely on adjustable measurement strength. We derive the analytical expressions for the average fidelity and the success probability of our teleportation protocols. The results show that the pre-flipping operator in the PF protocol effectively enhances the average fidelity of quantum teleportation in a deterministic manner, while the fidelity of PF-EAM remains at constant 1, unaffected by the noise intensity in the channel. In addition, our PF-EAM achieves a higher success probability while maintaining the fidelity of 1 with respect to the existing unprotected teleportation protocol, protocol based on weak measurement and measurement reversal (WM-MR), and protocol based on quantum feed-forward control and environment-assisted measurement (QFFC-EAM) framework.
Pauli check sandwiching is an error detection scheme that protects quantum circuits by inserting pairs of parity checks and discarding runs that signal errors. However, each additional check introduces noise and exponentially increases sampling costs. To address these limitations, we propose Pauli check extrapolation (PCE), an error mitigation technique that obtains measured expectation values from circuits with different numbers of checks and, analogous to zero-noise extrapolation (ZNE), extrapolates to the “maximum check” limit—the theoretical number of checks required for unit fidelity. We test linear and exponential ansatzes, deriving the exponential form from the Markovian error model. Benchmarking PCE against ZNE on random Clifford circuits with simulated depolarizing noise shows PCE outperforming ZNE for larger circuits. On real IBM hardware, PCE achieves an accuracy of up to 99.2% (56.2% improvement over baseline), compared to ZNE’s 82% accuracy (29.1% improvement over baseline), for four-qubit circuits. To demonstrate a practical use case, we then apply PCE toward mitigating errors in classical shadow measurements. Our results show that PCE can achieve fidelities greater than the state-of-the-art robust shadow estimation, while significantly reducing the number of required samples by eliminating the need for a calibration procedure. We validate these findings on both fully connected topologies and simulated IBM hardware backends.
