arXiv subjects

Xiwen Zhang

Publications and source records attributed to Xiwen Zhang.

35 records · Page 2Linked to original sources

arXiv2022

Structural reconstruction and anisotropic conductance in $4f$-ferromagnetic monolayer

Two-dimensional magnets are promising for nanoscale spintronic applications. Currently, most available candidates are based on $3d$ transition metal compounds, with hexagonal or honeycomb lattice geometry. Here, a GdCl$_3$ monolayer with $4f$ moments is theoretically studied, which can be exfoliated from its existing bulk. Its orthorhombic structure and hendecahedral ion cages are unique in two-dimensional. Furthermore, a significant structural reconstruction is caused by the implantation of Li atoms into its interstitial position, which also lead to ferromagnetism via a double-exchange-like process. Its highly anisotropic conductance may be peculiarly useful for nanoelectronics.

cond-mat.mtrl-sci↗

arXiv2020

Deep Learning for Frame Error Prediction using a DARPA Spectrum Collaboration Challenge (SC2) Dataset

We demonstrate a first example for employing deep learning in predicting frame errors for a Collaborative Intelligent Radio Network (CIRN) using a dataset collected during participation in the final scrimmages of the DARPA SC2 challenge. Four scenarios are considered based on randomizing or fixing the strategy for bandwidth and channel allocation, and either training and testing with different links or using a pilot phase for each link to train the deep neural network. We also investigate the effect of latency constraints, and uncover interesting characteristics of the predictor over different Signal to Noise Ratio (SNR) ranges. The obtained insights open the door for implementing a deep-learning-based strategy that is scalable to large heterogeneous networks, generalizable to diverse wireless environments, and suitable for predicting frame error instances and rates within a congested shared spectrum.

eess.SP↗

arXiv2020

Approximate Query Service on Autonomous IoT Cameras

Elf is a runtime for an energy-constrained camera to continuously summarize video scenes as approximate object counts. Elf's novelty centers on planning the camera's count actions under energy constraint. (1) Elf explores the rich action space spanned by the number of sample image frames and the choice of per-frame object counters; it unifies errors from both sources into one single bounded error. (2) To decide count actions at run time, Elf employs a learning-based planner, jointly optimizing for past and future videos without delaying result materialization. Tested with more than 1,000 hours of videos and under realistic energy constraints, Elf continuously generates object counts within only 11% of the true counts on average. Alongside the counts, Elf presents narrow errors shown to be bounded and up to 3.4x smaller than competitive baselines. At a higher level, Elf makes a case for advancing the geographic frontier of video analytics.

cs.DB↗

arXiv2020

Efficient Training of Deep Classifiers for Wireless Source Identification using Test SNR Estimates

We study efficient deep learning training algorithms that process received wireless signals, if a test Signal to Noise Ratio (SNR) estimate is available. We focus on two tasks that facilitate source identification: 1- Identifying the modulation type, 2- Identifying the wireless technology and channel in the 2.4 GHz ISM band. For benchmarking, we rely on recent literature on testing deep learning algorithms against two well-known datasets. We first demonstrate that using training data corresponding only to the test SNR value leads to dramatic reductions in training time while incurring a small loss in average test accuracy, as it improves the accuracy for low SNR values. Further, we show that an erroneous test SNR estimate with a small positive offset is better for training than another having the same error magnitude with a negative offset. Secondly, we introduce a greedy training SNR Boosting algorithm that leads to uniform improvement in accuracy across all tested SNR values, while using a small subset of training SNR values at each test SNR. Finally, we demonstrate the potential of bootstrap aggregating (Bagging) based on training SNR values to improve generalization at low test SNR values with scarcity of training data.

cs.LG↗

arXiv2019

FeatureExplorer: Interactive Feature Selection and Exploration of Regression Models for Hyperspectral Images

Feature selection is used in machine learning to improve predictions, decrease computation time, reduce noise, and tune models based on limited sample data. In this article, we present FeatureExplorer, a visual analytics system that supports the dynamic evaluation of regression models and importance of feature subsets through the interactive selection of features in high-dimensional feature spaces typical of hyperspectral images. The interactive system allows users to iteratively refine and diagnose the model by selecting features based on their domain knowledge, interchangeable (correlated) features, feature importance, and the resulting model performance.

cs.HC↗

arXiv2019

Spectral control over $γ$-ray echo using a nuclear frequency comb system

Two kinds of spectral control over $γ$-ray echo using a nuclear frequency comb system are theoretically investigated. A nuclear frequency comb system is composed of multiple nuclear targets under magnetization (hyperfine splitting), mechanical motion (Doppler shift) or both, namely, moving and magnetized targets. In frequency domain the unperturbed single absorption line of $γ$-ray therefore splits into multiple lines with equal spacing and becomes a nuclear frequency comb structure. We introduce spectral shaping and dynamical splitting to the frequency comb structure respectively to optimize the use of a medium and to break the theoretical maximum of echo efficiency, i.e., 54\%. Spectral shaping scheme leads to the reduction of required sample resonant thickness for achieving high echo efficiency of especially a broadband input. Dynamical splitting method significantly advances the echo efficiency up to 67\% revealed by two equivalent nuclear frequency comb systems. We also show that using only few targets is enough to obtain good echo performance, which significantly eases the complexity of implementation. Our results extend quantum optics to 10keV regime and lay the foundation of the development of $γ$-ray memory.

quant-ph↗

arXiv2019

Deep Learning for Interference Identification: Band, Training SNR, and Sample Selection

We study the problem of interference source identification, through the lens of recognizing one of 15 different channels that belong to 3 different wireless technologies: Bluetooth, Zigbee, and WiFi. We employ deep learning algorithms trained on received samples taken from a 10 MHz band in the 2.4 GHz ISM Band. We obtain a classification accuracy of around 89.5% using any of four different deep neural network architectures: CNN, ResNet, CLDNN, and LSTM, which demonstrate the generality of the effectiveness of deep learning at the considered task. Interestingly, our proposed CNN architecture requires approximately 60% of the training time required by the state of the art while achieving slightly larger classification accuracy. We then focus on the CNN architecture and further optimize its training time while incurring minimal loss in classification accuracy using three different approaches: 1- Band Selection, where we only use samples belonging to the lower and uppermost 2 MHz bands, 2- SNR Selection, where we only use training samples belonging to a single SNR value, and 3- Sample Selection, where we try various sub-Nyquist sampling methods to select the subset of samples most relevant to the classification task. Our results confirm the feasibility of fast deep learning for wireless interference identification, by showing that the training time can be reduced by as much as 30x with minimal loss in accuracy.

eess.SP↗

arXiv2018

Nuclear quantum memory and time sequencing of a single $γ$ photon

A $γ$-ray-nuclear quantum interface is suggested as a new platform for quantum information processing, motivated by remarkable progresses in $γ$-ray quantum optics. The main advantages of a $γ$ photon over an optical photon lie in its almost perfect detectability and much tighter, potentially sub-angstrom, focusability. Nuclear ensembles hold important advantages over atomic ensembles in a unique combination of high nuclear density in bulk solids with narrow, lifetime-broadening Mössbauer transitions even at room temperature. This may lead to the densest long-lived quantum memories and the smallest size photon processors. Here we propose a technique for $γ$ photon quantum memory through a Doppler frequency comb, produced by a set of resonantly absorbing nuclear targets that move with different velocities. It provides a reliable storage, an on-demand generation, and a time sequencing of a single $γ$ photon. This scheme presents the first $γ$-photon-nuclear-ensemble interface opening a new direction of research in quantum information science.

quant-ph↗

arXiv2017

Crack detection in beam structures with a novel Laplace based Wavelet Finite Element method

Beam structure is one of the most widely used structures in mechanical engineering and civil engineering. Ultrasonic guided wave based crack identification is one of the most important and accepted approaches applied to detect unseen small flaws in structures. Numerical simulations of ultrasonic guided wave propagation have caught more and more attention due to the fast development of hardware and software in the last few years. From all the numerical simulation methods, wavelet based finite element method has been proved to be one of the most efficient methods due to its better spatial resolution, which means it needs fewer elements to get the same accuracy and it can improve the calculation cost significantly. However, it needs a very small time interval. Laplace transform can easily convert the time domain into a frequency domain and then revert it back to a time domain. Laplace transform has thus the advantage of finding better results with a very large time interval. which can save a lot of time cost. This paper will present an innovative method combining Laplace transform and the B-spline wavelet on interval (BSWI) finite element method. This novel method allows to get results with the same accuracy and with a significantly lower time cost, which would not only decrease the total number of elements in the structure but also increase the time integration interval. The numerical Laplace transform and BSWI finite element will be introduced. Moreover, this innovative method is applied to simulate the ultrasonic wave propagation in a beam structure in different materials. Numerical examples for crack identification in beam structures have been studied for verification.

cs.CE↗

arXiv2016

Quantum storage based on controllable frequency comb

We suggest an all-optical scheme for the storage, retrieval and processing of a single-photon wave packet through its off-resonant Raman interaction with a series of coherent control beams. These control beams, each with distinct carrier frequency, are distributed along the way of single-photon propagation, thus effectively forming a gradient absorption structure which can be controlled in various ways to achieve different single-photon processing functionalities. Such a controllable frequency comb is a hybrid of Raman, gradient echo memory (GEM) and atomic frequency comb (AFC) methods, therefore demonstrates many of their advantages all together in one.

quant-ph↗

arXiv2016

Exact solution of gradient echo memory and analytical treatment of gradient frequency comb

Gradient echo memory (GEM) stores and retrieves photon wave packet in forward direction with high efficiency and fidelity using photon-echo mechanism. It is an important technique for quantum memory applications. By breaking the continuity of the gradient absorption structure, the scheme becomes gradient frequency comb (GFC), which is a hybrid of GEM and atomic frequency comb (AFC). To elucidate the non-trivial field-atom evolution of gradient echo, we derive its exact analytical solution in a medium of arbitrary optical thickness subjecting to any linear gradient of transition frequency, and discuss its physical processes. Based on this solution, we further suggest and analysis two types of GFC, one with stepwise gradient and the other with discontinuous gradient.

quant-ph↗

arXiv2015

Wetting Transitions of Condensed Droplets on Superhydrophobic Surfaces with Two-Tier Roughness

Although realizing wetting transitions of droplets spontaneously on solid rough surfaces is quite challenging, it is becoming a key research topic in many practical applications which require highly efficient removal of liquid. We report wetting transitions of condensed droplets occurring spontaneously on pillared surfaces with two-tier roughness owing to excellent superhydrophobicity. The phenomenon results from further decreased Laplace pressure on the top side of the individual droplet when its size becomes comparable to the scale of the micropillars, which leads to a surprising robust spontaneous wetting transition, from valleys to tops of the pillars. A simple scaling law is derived theoretically, which demonstrates that the critical size of the droplet is determined by the space of the micropillars. For this reason, highly efficient removal of water benefits greatly from smaller micropillar space. Furthermore, three wetting transition modes exist, in which the in situ wetting behaviors are in good agreement with our quantitative theoretical analysis.

cond-mat.soft↗

arXiv2014

All optical quantum storage based on spatial chirp of the control field

We suggest an all-optical quantum memory scheme which is based on the off-resonant Raman interaction of a signal quantum field and a strong control field in a three-level atomic medium in the case, when the control field has a spatially varying frequency across the beam, called a spatial chirp. We show that the effect of such a spatial chirp is analogous to the effect of a controllable reversible inhomogeneous broadening (CRIB) of the atomic transition used in the gradient echo memory (GEM) scheme. However, the proposed scheme does not require temporal modulation of the control field or the atomic levels, and can be realized without additional electric or magnetic fields. It means that materials demonstrating neither linear Stark nor Zeeman effects can be used and/or materials which are placed in specific external fields remain undisturbed.

quant-ph↗

arXiv2014

Parametric generation of high frequency coherent light in negative index materials and materials with strong anomalous dispersion

We demonstrate the possibility of generation of coherent radiation with tunable frequencies higher than the frequency of the driving field $ν_{d}$ in a nonlinear medium utilizing the difference combination resonance that occurs when $ν_{d}$ matches the difference of the frequencies of the two generated fields $ω_{1}$ and $ω_{2}$. We find that such a resonance can appear in materials which have opposite signs of refractive index at $ω_{1}$ and $ω_{2}$. It can also occur in positive refractive index materials with strong anomalous dispersion if at one of the generated frequencies the group and phase velocities are opposite to each other. We show that the light amplification mechanism is equivalent to a combination resonance in a system of two coupled parametric oscillators with the opposite sign of masses. Such a mechanism holds promise for a new kind of light source that emits coherent radiation of tunable wavelengths by an optical parametric amplification process with the frequency higher than $ν_{d}$.

physics.optics↗

arXiv2013

Superradiant control of gamma-ray propagation by vibrating nuclear arrays

The collective nature of light interactions with atomic and nuclear ensembles yields the fascinating phenomena of superradiance and radiation trapping. We study the interaction of gamma rays with a coherently vibrating periodic array of two-level nuclei. Such nuclear motion can be generated, e.g., in ionic crystals illuminated by a strong driving optical laser field. We find that deflection of the incident gamma beam into the Bragg angle can be switched on and off by nuclear vibrations on a superradiant time scale determined by the collective nuclear frequency, which is of the order of terahertz. Namely, if the incident gamma wave is detuned from the nuclear transition by much larger frequency it passes through the static nuclear array. However, if the nuclei vibrate with the frequency of the gamma ray detuning then parametric resonance can yield energy transfer into the Bragg deflected beam on the superradiant time scale, which can be used for fast control of gamma rays.

quant-ph↗

arXiv2013

Condensation and jumping relay of droplets on lotus leaf

Dynamic behavior of micro water droplet condensed on a lotus leaf with two-tier roughness is studied. Under laboratory environment, the contact angle of the micro droplet on single micro papilla increases smoothly from 80 deg to 160 deg during the growth of condensed water. The best-known "self-clean" phenomenon, will be lost. A striking observation is the out-of-plane jumping relay of condensed droplets triggered by falling droplets, as well as its sustained speed obtained in continuous jumping relays, enhance the automatic removal of dropwise condensation without the help from any external force. The surface tension energy dissipation is the main reason controlling the critical size of jumping droplet and its onset velocity of rebounding.

physics.flu-dyn↗

arXiv2012

Quantum storage based on the control field angular scanning

Continuous change of the propagation direction of a classical control field in the process of its off-resonant Raman interaction with a weak signal field in a three-level atomic medium is suggested for quantum storage of a single-photon wave packet. It is shown that due to phase matching condition such an angular control allows one to reversibly map the single-photon wave packet to the Raman spatial coherence grating. Thus, quantum storage and retrieval can be realized without using inhomogeneous broadening of the atomic transitions or manipulating the amplitude of the control field. Under some conditions the proposed scheme proves to be mathematically analogous to the quantum storage scheme based on controlled reversible inhomogeneous broadening of the atomic states.

quant-ph↗