Search

Multiple federated learning (FL) methods are proposed for traffic flow forecasting (TFF) to avoid heavy-transmission and privacy-leaking concerns resulting from the disclosure of raw data in centralized methods. However, these FL methods adopt offline learning which may yield subpar performance, when concept drift occurs, i.e., distributions of historical and future data vary. Online learning can detect concept drift during model training, thus more applicable to TFF. Nevertheless, the existing federated online learning method for TFF fails to efficiently solve the concept drift problem and causes tremendous computing and communication overhead. Therefore, we propose a novel method named Resource-Efficient Federated Online Learning (REFOL) for TFF, which guarantees prediction performance in a communication-lightweight and computation-efficient way. Specifically, we design a data-driven client participation mechanism to detect the occurrence of concept drift and determine clients' participation necessity. Subsequently, we propose an adaptive online optimization strategy, which guarantees prediction performance and meanwhile avoids meaningless model updates. Then, a graph convolution-based model aggregation mechanism is designed, aiming to assess participants' contribution based on spatial correlation without importing extra communication and computing consumption on clients. Finally, we conduct extensive experiments on real-world datasets to demonstrate the superiority of REFOL in terms of prediction improvement and resource economization.

How can the stability and efficiency of Artificial Neural Networks (ANNs) be ensured through a systematic analysis method? This paper seeks to address that query. While numerous factors can influence the learning process of ANNs, utilizing knowledge from control systems allows us to analyze its system function and simulate system responses. Although the complexity of most ANNs is extremely high, we still can analyze each factor (e.g., optimiser, hyperparameters) by simulating their system response. This new method also can potentially benefit the development of new optimiser and learning system, especially when discerning which components adversely affect ANNs. Controlling ANNs can benefit from the design of optimiser and learning system, as (1) all optimisers act as controllers, (2) all learning systems operate as control systems with inputs and outputs, and (3) the optimiser should match the learning system. Please find codes: \url{https://github.com/RandomUserName2023/Control-ANNs}., Comment: 23 pages

Interactive volume visualization using a mixed reality (MR) system helps provide users with an intuitive spatial perception of volumetric data. Due to sophisticated requirements of user interaction and vision when using MR head-mounted display (HMD) devices, the conflict between the realisticness and efficiency of direct volume rendering (DVR) is yet to be resolved. In this paper, a new MR visualization framework that supports interactive realistic DVR is proposed. An efficient illumination estimation method is used to identify the high dynamic range (HDR) environment illumination captured using a panorama camera. To improve the visual quality of Monte Carlo-based DVR, a new spatio-temporal denoising algorithm is designed. Based on a reprojection strategy, it makes full use of temporal coherence between adjacent frames and spatial coherence between the two screens of an HMD to optimize MR rendering quality. Several MR development modules are also developed for related devices to efficiently and stably display the DVR results in an MR HMD. Experimental results demonstrate that our framework can better support immersive and intuitive user perception during MR viewing than existing MR solutions., Comment: 6 pages, 6 figures

Generation of single photons carrying spin and orbital angular momenta (SAM and OAM) opens enticing perspectives for exploiting multiple degrees of freedom for high-dimensional quantum systems. However, on-chip generation of single photons encoded with single-mode SAM-OAM states has been a major challenge. Here, by utilizing carefully designed anisotropic nanodimers fabricated atop a substrate, supporting surface plasmon polariton (SPP) propagation, and accurately positioned around a quantum emitter (QE), we enable nonradiative QE-SPP coupling and the SPP outcoupling into free-space propagating radiation featuring the designed SAM and OAM. We demonstrate on-chip room-temperature generation of well-collimated (divergence < 7.5 degrees) circularly polarized (chirality > 0.97) single-mode vortex beams with different topological charges (l = 0, 1, and 2) and high single-photon purity, g(0) < 0.15. The developed approach can straightforwardly be extended to produce multiple, differently polarized, single-mode single-photon radiation channels, and enable thereby realization of high-dimensional quantum sources for advanced quantum photonic technologies., Comment: 15 pages and 4 figures in the main text; 24 pages and 19 figures in supplementary information

Ultracompact chip-integrated single-photon sources of collimated beams with polarizationencoded states are crucial for integrated quantum technologies. However, most of currently available single-photon sources rely on external bulky optical components to shape the polarization and phase front of emitted photon beams. Efficient integration of quantum emitters with beam shaping and polarization encoding functionalities remains so far elusive. Here, we present ultracompact single-photon sources of linearly polarized vortex beams based on chip-integrated quantum emitter-coupled metasurfaces, which are meticulously designed by fully exploiting the potential of nanobrick arrayed metasurfaces. We first demonstrate on-chip single-photon generation of high-purity linearly polarized vortex beams with prescribed topological charges of -1, 0, and +1. We further realize multiplexing of single-photon emission channels with orthogonal linear polarizations carrying different topological charges and demonstrate their entanglement. Our work illustrates the potential and feasibility of ultracompact quantum emitter-coupled metasurfaces as a new quantum optics platform for realizing chip-integrated high-dimensional single-photon sources., Comment: 17 pages and 4 figures in the main text; 22 pages and 14 figures in supplementary information

In the federated learning scenario, geographically distributed clients collaboratively train a global model. Data heterogeneity among clients significantly results in inconsistent model updates, which evidently slow down model convergence. To alleviate this issue, many methods employ regularization terms to narrow the discrepancy between client-side local models and the server-side global model. However, these methods impose limitations on the ability to explore superior local models and ignore the valuable information in historical models. Besides, although the up-to-date representation method simultaneously concerns the global and historical local models, it suffers from unbearable computation cost. To accelerate convergence with low resource consumption, we innovatively propose a model regularization method named FedTrip, which is designed to restrict global-local divergence and decrease current-historical correlation for alleviating the negative effects derived from data heterogeneity. FedTrip helps the current local model to be close to the global model while keeping away from historical local models, which contributes to guaranteeing the consistency of local updates among clients and efficiently exploring superior local models with negligible additional computation cost on attaching operations. Empirically, we demonstrate the superiority of FedTrip via extensive evaluations. To achieve the target accuracy, FedTrip outperforms the state-of-the-art baselines in terms of significantly reducing the total overhead of client-server communication and local computation., Comment: 11 pages, to be published in the 2023 IEEE International Parallel and Distributed Processing Symposium (IPDPS)

As a nontrivial solid state of matter, the glassy-crystal state embraces physical features of both crystalline and amorphous solids, where a long-range ordered periodic structure formed by the mass centers of constituent molecules accommodates orientational glasses. Here, we discover and validate a glassy-crystal state in 2-amino-2-methyl-1,3-propanediol (AMP, C4H11NO2) by neutron scattering and complementary broadband dielectric spectroscopy (BDS) measurements. The freezing process of the dynamic orientational disorder is manifested at relaxation times well described by the Vogel-Fulcher-Tammann (VFT) law and the strongly frequency-dependent freezing temperature ranging from around 225 K at 0.1 Hz to above room temperature in the GHz region. At room temperature, the supercooled state is extremely sensitive to pressure such that a few MPa pressure can induce crystallization to the ordered crystal state, eventually leading to a temperature increase by 48 K within 20 s, a significant reduction of visible light transmittance from about 95% to a few percentages, and a remarkable decrease of electrical conductivity by three orders of magnitude. These ultrasensitive baroresponsivities might find their applications in low-grade waste heat recycling, pressure sensors and non-volatile memory devices. It is expected that glassy crystals serve as an emerging platform for exploiting exotic states of matter and the associated fantastic applications., Comment: Totally 13 figures and 1 table

Emergent phenomena at heterointerfaces are directly associated with the bonding geometry of adjacent layers. Effective control of accessible parameters, such as the bond length and bonding angles, offers an elegant method to tailor competing energies of the electronic and magnetic ground states. In this study, we construct unit thick syntactic layers of cobaltites within a strongly tilted octahedral matrix via atomically precise synthesis. The octahedral tilt patterns of adjacent layers propagate into cobaltites, leading to a continuation of octahedral tilting while maintaining significant misfit tensile strain. These effects induce severe rumpling within an atomic plane of neighboring layers triggers the electronic reconstruction between the splitting orbitals. First-principles calculations reveal that the cobalt ions transits to a higher spin state level upon octahedral tilting, resulting in robust ferromagnetism in ultrathin cobaltites. This work demonstrates a design methodology for fine-tuning the lattice and spin degrees of freedom in correlated quantum heterostructures by exploiting epitaxial geometric engineering., Comment: 18 pages, 5 figures

A new conservative finite element solver for the three-dimensional steady magnetohydrodynamic (MHD) kinematics equations is presented.The solver utilizes magnetic vector potential and current density as solution variables, which are discretized by H(curl)-conforming edge-element and H(div)-conforming face element respectively. As a result, the divergence-free constraints of discrete current density and magnetic induction are both satisfied. Moreover the solutions also preserve the total magnetic helicity. The generated linear algebraic equation is a typical dual saddle-point problem that is ill-conditioned and indefinite. To efficiently solve it, we develop a block preconditioner based on constraint preconditioning framework and devise a preconditioned FGMRES solver. Numerical experiments verify the conservative properties, the convergence rate of the discrete solutions and the robustness of the preconditioner., Comment: 13 pages. arXiv admin note: text overlap with arXiv:1712.08922

Defects ubiquitously exist in crystal materials and usually exhibit a very different nature than the bulk matrix, and hence, their presence can have significant impacts on the properties of devices. Although it is well accepted that the properties of defects are determined by their unique atomic environments, the precise knowledge of such relationships is far from clear for most oxides due to the complexity of defects and difficulties in characterization. Here, we fabricate a 36.8{\deg} SrRuO3 grain boundary of which the transport measurements show a spin-valve magnetoresistance. We identify its atomic arrangement, including oxygen, using scanning transmission electron microscopy and spectroscopy. Based on the as-obtained atomic structure, the density functional theory calculations suggest that the spin-valve magnetoresistance is because of the dramatically reduced magnetic moments at the boundary. The ability to manipulate magnetic properties at the nanometer scale via defect control allows new strategies to design magnetic/electronic devices with low-dimensional magnetic order., Comment: 26 pages,4 figures

Edge structures are highly relevant to the electronic, magnetic and catalytic properties of two-dimensional (2D) transition metal dichalcogenides (TMDs) and their one dimensional (1D) counterpart, i.e., nanoribbons, which should be precisely tailored for the desirable applications. In this work, we report the formation of novel Mo6S6 nanowire (NW) terminated edges in a monolayer molybdenum disulfide (MoS2) via an e-beam irradiation process combined with high temperature heating in a scanning transmission electron microscope (STEM). Atomic structures of NW terminated edges and the dynamic formation process were observed experimentally. Further analysis shows that NW terminated edge could form on both Mo-zigzag (ZZ) edge and S-ZZ edge which can exhibit even higher stability superior to the pristine zigzag (ZZ) and armchair (AC) edge. In addition, the analogous edge structures can be also formed in MoS2 nanoribbon and other TMDs material such as MoxW1-xSe2. We believe that the presence of these novel edge structures in 2D and 1D TMD materials may provide novel properties and new opportunities for their versatile applications including catalytic, spintronic and electronic devices., Comment: 10 pages,4 figures

Dislocations in perovskite oxides have important impacts on their physical and chemical properties, which are determined by their unique atomic environments. In the present study, the structure of dislocations in a 10{\deg} low-angle grain boundary of SrTiO3 (STO) is characterized by spherical aberration corrected scanning transmission electron microscopy (Cs-STEM) and spectroscopy. In contrast to previous studies, the deficiency instead of enrichment of titanium (Ti) is observed at the dislocation cores mainly due to the Sr substitution and under occupancy of Ti. The presence of oxygen vacancies and partially reduced Ti are also detected at the Ti deficient dislocations cores. These findings indicate the atomic structure of dislocations can be very different even they have the same Burges vectors. Controllable elemental segregation in the dislocations and grain boundaries via bicrystal engineering should be very useful for design of devices with novel functions., Comment: 14 pages, 4 figures