기존 양자 패킷 스위칭 방식에서는 양자 네트워크를 확장하기 위해 고전 헤더와 양자 페이로드 간의 시간 정렬을 유지해야 하지만, 제어 평면의 처리 지연과 지터로 인해 이를 안정적으로 유지하기 어렵다. 본 연구에서는 이러한 문제를 해결하기 위해 라우팅 라벨을 인코딩된 양자 페이로드의 신드롬 구조에 삽입하는 Syndrome-as-Header(SAH) 양자 라벨 스위칭 구조를 제안한다. SAH는 UEI(Uncorrectable Error Injection)를 이용하여 플로우 라벨을 기준 신드롬에 매핑하며, 코어 라우터는 논리 페이로드를 측정하거나 디코딩하지 않고 신드롬 헤더를 추출하여 채널 오류 성분을 제거하고 라벨을 교체할 수 있다.

The Tail Assignment problem, a central challenge in airline scheduling, involves assigning aircraft to flights while minimizing operational costs. Building on prior work that formulates this problem as an Exact Cover instance and solves it using QAOA, we extend the approach by applying Recursive QAOA (RQAOA) - a variant that iteratively reduces the problem size by fixing variables based on intermediate solutions. Our simulation results show that increasing the number of iterations of level-1 RQAOA yields better performance than increasing the level of QAOA. Furthermore, RQAOA delivers superior solution quality with reduced quantum resource requirements, as it relies on repeated shallow circuits rather than deeper quantum evolutions. This approach achieves a high success probability in identifying the feasible solution and suggests potential performance improvements on near-term quantum hardware.  

Multipartite quantum entanglement plays a crucial role in the emergence of different quantum phases and their transitions in quantum many-body systems. It is of general interest to know what sort of analysis on quantum entanglement can bring us a profound insight to understand the rich dynamics of quantum many-body systems. In this work we study the characteristics of quantum entanglement in relation to -ordered phases emerging under a varied strength of 1-dim Rydberg interaction. We propose an approach based on the structure of pair-wise entanglement across the Rydberg chain using two-qubit concurrence as an entanglement measure. We define an entanglement-structure factor via Fourier analysis of total concurrence at each site and address  phase transitions in comparison with the conventional order-parameter based on local density, i.e. magnetization. We also discuss how the required two-qubit concurrence can be measured in analog Rydberg atom arrays using site-selective erasure and parametrized laser pulses. Our investigation suggests that an entanglement-structure-based approach can provide a powerful tool in analyzing symmetry-breaking in quantum phase transitions. 

Simulating quantum dynamics in finite-sized open systems, such as Rydberg atom chains, often suffers from non-physical reflections at artificial boundaries that alter the internal bulk state dynamics. In this work, we propose a learned boundary closure framework to suppress finite-size boundary reflections in one-dimensional Rydberg atom chains operating in the blockade regime (PXP model). By parameterizing loss rates and local detunings on the outermost two atoms, our model dynamically tracks and absorbs incoming wavepackets through real-time feedback on local observables. Trained against a 96-site Time-Evolving Block Decimation (TEBD) benchmark, the closed-loop policy reduces density errors by 43% to 79% across various system sizes (L = 8, 10, 12) compared to conventional hard-wall boundaries. Furthermore, the learned closure demonstrates strong generalization to unseen detunings outside its training range. Remarkably, achieving this accuracy level on a 10-site system using our boundary closure outperforms a 12-site system with hard-wall boundaries, saving significant computational memory without introducing an error floor. 

본 연구는 NOAA Global Drifter Program의 시간별 품질관리 자료를 활용하여 고등학생이 실제 관측자료로 정량적 벡터 해류도를 제작·해석하는 STEAM 기반 탐구 프로그램을 개발하고, 소규모 파일럿 적용에서 나타난 학습 경험을 분석하였다. 경기도 소재 A고등학교 2학년 학생 5명이 대학 연계 진로캠프에 참여하였다. 프로그램은 해류와 표층 뜰개 개념(Science), 위성추적 자료·ERDDAP·Google Colab 활용(Technology), 자료 처리 절차와 코드 수정(Engineering), 화살표·색·범례 및 포스터 설계(Arts), ve·vn 벡터·격자 평균·통계(Mathematics)를 하나의 문제 해결 흐름으로 연결하였다. 학생용 Colab의 주 처리에서는 수동으로 내려받은 2012–2021년 여섯 해역 CSV 6,390,812행을 점검한 뒤 2° 격자별 mean_ve와 mean_vn으로 평균하고, 관측 200개·뜰개 3개 이상인 559개 격자를 유지하였다. 교사 수합용 Colab은 동일 길이, 가시성 보정 제곱근 길이, 96시간 선형 길이를 비교하였다. 24개 자기인식 문항 평균은 2.01에서 4.86으로 증가했으며, 학생들은 최종 포스터에서 해류 변화와 쓰레기 집적의 관계까지도 논의하였다.

Sample-based quantum diagonalization (SQD) identifies important electronic configurations through quantum-compatible sampling and performs classical diagonalization in the resulting subspace. We extend this approach to a 19-point reactive geometry path using a small number of directly calculated anchor states. For each target geometry, configurations are sampled independently from the two nearest anchors, refined using the target Hamiltonian, and supplemented through residual leakage sampling. The remaining residual is then used to select the next anchor without access to the exact ground state. In an ideal statevector-emulator benchmark, seven anchors replaced nineteen pointwise state acquisitions, and  adaptive anchor placement reduced the worst-target error by 42.9% compared with  fixed placement. These results demonstrate a strategy for concentrating electronic-structure effort in difficult regions of a reaction path, without claiming quantum hardware execution or quantum advantage.

Quantifying information measures in a quantum system is a foundational task in quantum information theory, with applications ranging from quantum communication to resource theories. However, evaluating such measures directly often requires intractable optimizations that pose significant computational challenges. In this work, we propose a fully operational framework to estimate the quantum umlaut information utilizing a variational quantum generative adversarial network (QGAN). Our approach leverages recent advances in quantum machine learning to overcome the computational bottlenecks of conventional methods, providing a practical and scalable route to estimating quantum information quantities. Specifically, we introduce the measured quantum umlaut information and transform it into a min-max objective function suitable for adversarial optimization, in which the generator and discriminator are realized as parameterized quantum circuits trained in a competitive manner. Crucially, we provide a strict mathematical guarantee, proving that the measured quantum umlaut information serves as an epsilon-approximate of the exact quantum umlaut information, with explicit bounds quantifying the approximation error. Our results establish a principled framework for the variational estimation of quantum information measures, opening promising directions for studying quantum correlations, communication tasks, and resource-theoretic problems within a unified adversarial learning paradigm.

This presentation explores two examples of quantum computing, ranging from quantum information processing to structural mechanics. First, I introduce an online quantum anomaly detection problem and explain how quantum memory can provide a significant advantage over classical memory in sequential state discrimination.

I then present a VQE-based approach to structural mechanics, where FEM-derived mass and stiffness matrices are transformed into an eigenvalue problem and solved using the Variational Quantum Eigensolver. Results for truss, beam, and continuum models are used to discuss both the potential of quantum algorithms for mechanical engineering and the current limitations of quantum hardware.

The ionosphere is a partially ionized region of the upper atmosphere in which free electrons and ions form a dispersive plasma, producing frequency-dependent delays in Global Navigation Satellite System (GNSS) signals and directly affecting precise positioning. Total electron content (TEC), defined as the electron density integrated along a signal propagation path, is therefore both an essential GNSS correction parameter and a sensitive indicator of ionospheric variability. TEC observations have been used not only to monitor solar flares and geomagnetic disturbances but also to investigate ionospheric responses to earthquakes, volcanic eruptions, and severe atmospheric systems. Tropical cyclones can perturb the upper atmosphere through vertical coupling processes. However, many previous studies have relied on vertical TEC (VTEC), a column-integrated quantity that cannot resolve altitude-dependent changes in electron density. In this study, we investigate three-dimensional ionospheric perturbations before, during, and after the passage of Typhoon Khanun over the Korean Peninsula in 2023 using regional multi-GNSS ionospheric tomography. Slant TEC (STEC) is derived from dual-frequency GNSS observations along satellite–receiver paths selected using a 30° elevation-angle cutoff. The International Reference Ionosphere 2020 (IRI-2020) model is used as the initial background ionosphere, and the multiplicative algebraic reconstruction technique (MART) is applied to iteratively adjust the three-dimensional electron-density field along the GNSS ray paths. The reconstruction minimizes the discrepancy between the observed STEC and the STEC predicted from the reconstructed electron-density field. We examine whether the tomographically reconstructed electron-density field and derived VTEC can resolve ionospheric perturbations associated with the passage of the typhoon. Our results demonstrate that regional ionospheric tomography provides vertical information beyond that available from conventional VTEC monitoring and has the potential to support near-real-time characterization of ionospheric responses to severe atmospheric events.

This presentation explores two examples of quantum computing, ranging from quantum information processing to structural mechanics. First, I introduce an online quantum anomaly detection problem and explain how quantum memory can provide a significant advantage over classical memory in sequential state discrimination.

I then present a VQE-based approach to structural mechanics, where FEM-derived mass and stiffness matrices are transformed into an eigenvalue problem and solved using the Variational Quantum Eigensolver. Results for truss, beam, and continuum models are used to discuss both the potential of quantum algorithms for mechanical engineering and the current limitations of quantum hardware.

This presentation explores two examples of quantum computing, ranging from quantum information processing to structural mechanics. First, I introduce an online quantum anomaly detection problem and explain how quantum memory can provide a significant advantage over classical memory in sequential state discrimination.

I then present a VQE-based approach to structural mechanics, where FEM-derived mass and stiffness matrices are transformed into an eigenvalue problem and solved using the Variational Quantum Eigensolver. Results for truss, beam, and continuum models are used to discuss both the potential of quantum algorithms for mechanical engineering and the current limitations of quantum hardware.