Published papers

Entanglement generation between distant nodes is a fundamental process in distributed quantum information processing. Qudits, high-dimensional quantum states, are promising candidates for enhancing entanglement distribution capabilities. However, the success probability of qudit entanglement distribution using Bell measurements is typically lower than that of conventional qubit-based protocols. In this paper, we propose a novel entanglement swapping protocol specifically designed for qutrits (three-dimensional quantum states). Our protocol employs photon-number encoding combined with an additional mode basis, such as polarization, effectively increasing the success probability. We show that the proposed qutrit protocol can achieve higher entanglement generation rates compared to conventional two-photon detection-based qubit protocols, especially under conditions of limited photon generation probability. Furthermore, we evaluate our protocol under realistic experimental imperfections, including photon loss and threshold detection, and show that it achieves high fidelity with probabilistic photon sources.

Entanglement swapping at telecom wavelengths is at the heart of quantum networking in optical fiber infrastructures. Although entanglement swapping has been demonstrated experimentally so far using various types of entangled photon sources both in near-infrared and telecom wavelength regions, the rate of swapping operation has been too low to be applied to practical quantum protocols, due to limited efficiency of entangled photon sources and photon detectors. Here we demonstrate drastic improvement of the efficiency at telecom wavelength by using two ultra-bright entangled photon sources and four highly efficient superconducting nanowire single photon detectors. We have attained a four-fold coincidence count rate of 108 counts per second, which is three orders higher than the previous experiments at telecom wavelengths. A raw (net) visibility in a Hong-Ou-Mandel interference between the two independent entangled sources was 73.3 ± 1.0% (85.1 ± 0.8%). We performed the teleportation and entanglement swapping, and obtained a fidelity of 76.3% in the swapping test. Our results on the coincidence count rates are comparable with the ones ever recorded in teleportation/swapping and multi-photon entanglement generation experiments at around 800â€...nm wavelengths. Our setup opens the way to practical implementation of device-independent quantum key distribution and its distance extension by the entanglement swapping as well as multi-photon entangled state generation in telecom band infrastructures with both space and fiber links.

The demonstration of superadditive quantum-coding gain experiment was analyzed. A design strategy of quantum-collective decoding in physical quantum circuits was emphasized. The superadditive quantum coding became essential in the region where the noncommutativity of signal states was the main ambiguity among signals. It was stated that only quantum decoding can be the core technology to achieve the ultimate performance in future long-haul optical communications.

We investigate the possibility of implementing a given projection measurement using linear optics and arbitrarily fast feedforward based on the continuous detection of photons. In particular, we systematically derive the so-called Dolinar scheme that achieves the minimum-error discrimination of binary coherent states. Moreover, we show that the Dolinar-type approach can also be applied to projection measurements in the regime of photonic-qubit signals. Our results demonstrate that for implementing a projection measurement with linear optics, in principle, unit success probability may be approached even without the use of expensive entangled auxiliary states, as they are needed in all known (near-)deterministic linear-optics proposals. © 2005 The American Physical Society.

Polarization entanglement is widely used in optical quantum information processing due to its compatibility with standard optical components. On the other hand, it is known that polarization entanglement is susceptible to the loss, more precisely, its transmission rate in a lossy channel is limited by the scaling of O(η), where η is a transmittance of the channel. Here, we experimentally demonstrate that this rate-loss scaling limit can be overcome by a relatively simple protocol. This is possible by integrating the idea of the polarizaion-photon-number hybrid entanglement and the single-click entanglement swapping. We demonstrate square root improvement of the rate-loss scaling from the conventional approaches and achieve the fidelity of 0.843 for the distributed polarization entangled photon pairs. This improvement in the rate-loss scaling is equivalent to that achieved by 1-hop quantum repeater node. Our result paves a way to build a near-future quantum network and its applications.

We investigate the minimum achievable bit error rate of the discrimination of binary coherent states with the help of arbitrary ancillary states. We adopt homodyne measurement with a common phase of the local oscillator and classical feedforward control. After one ancillary state is measured, its outcome is referred to the preparation of the next ancillary state and the tuning of the next mixing with the signal. It is shown that the minimum bit error rate of the system is invariant under the following operations: feedforward control, deformations, and introduction of any ancillary state. We also discuss the possible generalization of the homodyne detection scheme. © 2010 The American Physical Society.

A theoretical analysis is given for a two-photon subtraction from a continuous-wave squeezed vacuum with finite time separation between two detection events. In the cw photon subtraction process, generated states are inevitably described by temporal multimode states. Our approach is based on analytical formulas that are mathematically simple and provide an intuitive understanding of the multimode structure of the system. We show that, in our process, the photon-subtracted squeezed vacuum is generated in two temporal modes and one of these modes acts as an ancillary mode to make the other one a large-amplitude coherent-state superposition. © 2008 The American Physical Society.

Open-Domain Multi-Document Summarization (ODMDS) is the task of generating summaries from large document collections in response to user queries. This task is crucial for efficiently addressing diverse information needs from users. Traditional retrieve-then-summarize approaches fall short for open-ended queries in ODMDS tasks. These queries often require broader context than initially retrieved passages provide, making it challenging to retrieve all relevant information in a single search. While iterative retrieval methods has been explored for multi-hop question answering (MQA), it's impractical for ODMDS due to high latency from repeated LLM inference. Accordingly, we propose LightPAL, a lightweight passage retrieval method for ODMDS. LightPAL leverages an LLM to pre-construct a graph representing passage relationships, then employs random walk during retrieval, avoiding iterative LLM inference. Experiments demonstrate that LightPAL outperforms naive sparse and pre-trained dense retrievers in both retrieval and summarization metrics, while achieving higher efficiency compared to iterative MQA approaches.

Device-independent quantum key distribution (DI-QKD) enables two remote parties to share an information-theoretically secure key without any assumptions on the inner workings of the devices used. Device-independent conference key agreement (DI-CKA) is multipartite DI-QKD in which more than two parties share a common secure key. The performance of DI-CKA, however, is strictly limited because of its susceptibility to losses due to, e.g., imperfect detection efficiency and channel transmission. Here, we propose a DI-CKA protocol which reduces this limitation by using a heralding scheme to distribute multipartite entanglement. We analyze key rates of our protocol for two different measurement scenarios, and we show that our protocol outperforms a previous DI-CKA protocol even with an experimentally feasible measurement.

Device-independent quantum key distribution (DI-QKD) enables information-theoretically secure key exchange between remote parties without any assumptions on the internal workings of the devices used for its implementation. However, its practical deployment remains severely constrained by the need for loophole-free Bell inequality violations, which are highly susceptible to losses and detection efficiencies. In this paper, we propose two long-distance DI-QKD protocols based on a heralding scheme using single-photon interference. Our protocols consist of only standard quantum optics tools such as two-mode squeezed states, displacement operations and on-off detectors, making them experimentally accessible. To further enhance robustness against realistic imperfections, we integrate a classical noisy preprocessing technique during post-processing. We calculate key rates of the protocols by numerical optimization and show the supremacy of this implementation over existing protocols in terms of communication distances.

A novel low-threshold amplification scheme for bichromatically coupling three upper states including a metastable state is investigated. Based on model calculations, it is possible to amplify coherent vacuum-ultraviolet emission in Kr gases without population inversion at the probed transition. In this system, therefore, media that are the same as those used for lasing (e.g., Kr) can be used as the incoherent pumping source. The low-threshold amplification is attained when the spontaneous emission at the probed transition is modulated through population trapping in the upper states by two intense coupling lasers. The possibility of continuous amplification of vacuum-ultraviolet emission in this model is discussed. © 1998 Optical Society of America.

We propose and analyze a novel low-threshold amplification scheme by bichromatically coupling three upper states including a metastable state. We numerically show that it is possible to amplify VUV emission in Kr gases without population inversion at the probe transition. It is shown that the origin of low-threshold gain is the population trapping of upper states dressed by two intense coupling lasers, and there is a possibility of forming large population inversion in optically dressed systems. ©2004 Copyright SPIE - The International Society for Optical Engineering.

On the current communication scheme, Gaussian coherent state signals play an important role. To make use of a potential of coherent state signals going beyond the shot noise limit, not only Gaussian operations, but also non-Gaussian operations, described as the third or higher order interactions with respect to the electric field amplitude, must essentially be applied. In this manuscript, we discuss our recent results on the enhancement of entanglement and quantum signal descrimination via the measurement-induced non-Gaussian operation with the photon detector and linear optics.

The dependence of the initial pulse waveform on spectral filtering squeezing and the optimization of pulses to obtain the further squeezing are numerically studied. Calculations are performed by solving the quantum nonlinear Schrodinger equation (QNLSE) in the positive P representation. The optimum pulse shaping can induce more effective quantum correlation between the spectral components of the pulses though the process of pulse shaping is entirely classical.

We study a qubit-oscillator system, with a time-dependent coupling coefficient, and present a fast scheme for generating entangled Schrödinger-cat states with large mean photon numbers and also a scheme that protects the cat states against dephasing caused by the nonlinearity in the system. We focus on the case where the qubit frequency is small compared to the oscillator frequency. We first present the exact quantum state evolution in the limit of infinitesimal qubit frequency. We then analyze the first-order effect of the nonzero qubit frequency. Our scheme works for a wide range of coupling strength values, including the recently achieved deep-strong-coupling regime.

The probabilities of vibronic transitions in molecules are referred to as Franck-Condon factors (FCFs). Although several approaches for calculating FCFs have been developed, such calculations are still challenging. Recently it was shown that there exists a correspondence between the problem of calculating FCFs and boson sampling. However, if the output photon number distribution of boson sampling is sparse, then it can be classically simulated. Exploiting these results, we develop a method to approximately reconstruct the distribution of FCFs of certain molecules. We demonstrate its proof of concept by applying it to formic acid and thymine at 0 K. In our method, we first obtain the marginal photon number distributions for pairs of modes of a Gaussian state associated with the molecular transition. We then apply a compressive sensing method called polynomial-time matching pursuit to recover FCFs.