arXiv subjects

Xiao Li

Publications and source records attributed to Xiao Li.

At least 505 records · Page 28Linked to original sources

arXiv2015

An unconditionally energy stable finite difference scheme for a stochastic Cahn-Hilliard equation

In this work, the MMC-TDGL equation, a stochastic Cahn-Hilliard equation is solved numerically by using the finite difference method in combination with a convex splitting technique of the energy functional. For the non-stochastic case, we develop an unconditionally energy stable difference scheme which is proved to be uniquely solvable. For the stochastic case, by adopting the same splitting of the energy functional, we construct a similar and uniquely solvable difference scheme with the discretized stochastic term. The resulted schemes are nonlinear and solved by Newton iteration. For the long time simulation, an adaptive time stepping strategy is developed based on both first- and second-order derivatives of the energy. Numerical experiments are carried out to verify the energy stability, the efficiency of the adaptive time stepping and the effect of the stochastic term.

math.NA↗

arXiv2015

Mixed and missing data: a unified treatment with latent graphical models

We propose to learn latent graphical models when data have mixed variables and missing values. This model could be used for further data analysis, including regression, classification, ranking etc. It also could be used for imputing missing values. We specify a latent Gaussian model for the data, where the categorical variables are generated by discretizing an unobserved variable and the latent variables are multivariate Gaussian. The observed data consists of two parts: observed Gaussian variables and observed categorical variables, where the latter part is considered as partially missing Gaussian variables. We use the Expectation-Maximization algorithm to fit the model. To prevent overfitting we use sparse inverse covariance estimation to obtain sparse estimate of the latent covariance matrix, equivalently, the graphical model. The fitted model then could be used for problems including re- gression, classification and ranking. Such an approach is applied to a medical data set where our method outperforms the state-of-the-art methods. Simulation studies and real data results suggest that our proposed model performs better than random forest in terms of prediction error when the model is correctly specified, and is a better imputation method than hot deck imputation even if the model is not correctly specified.

stat.ME↗

arXiv2015

On the Lagrangian angle and the Kähler angle of immersed surfaces in the complex plane $\bbc^2$

In this paper, we discuss the Lagrangian angle and the Kähler angle of immersed surfaces in $\mathbb C^2$. Firstly, we provide an extension of Lagrangian angle, Maslov form and Maslov class to more general surfaces in $\mathbb C^2$ than Lagrangian surfaces, and then naturally extend a theorem by J.-M. Morvan to surfaces of constant Kähler angle, together with an application showing that the Maslov class of a compact self-shrinker surface with constant Kähler angle is generally non-vanishing. Secondly, we obtain two pinching results for the Kähler angle which imply rigidity theorems of self-shrinkers with Kähler angle under the condition that $\int_M |h|^2e^{-\frac{|x|^2}{2}}dV_M<\infty$, where $h$ and $x$ denote, respectively, the second fundamental form and the position vector of the surface.

math.DG↗

arXiv2015

SPRIGHT: A Fast and Robust Framework for Sparse Walsh-Hadamard Transform

We consider the problem of computing the Walsh-Hadamard Transform (WHT) of some $N$-length input vector in the presence of noise, where the $N$-point Walsh spectrum is $K$-sparse with $K = {O}(N^δ)$ scaling sub-linearly in the input dimension $N$ for some $0<δ<1$. Over the past decade, there has been a resurgence in research related to the computation of Discrete Fourier Transform (DFT) for some length-$N$ input signal that has a $K$-sparse Fourier spectrum. In particular, through a sparse-graph code design, our earlier work on the Fast Fourier Aliasing-based Sparse Transform (FFAST) algorithm computes the $K$-sparse DFT in time ${O}(K\log K)$ by taking ${O}(K)$ noiseless samples. Inspired by the coding-theoretic design framework, Scheibler et al. proposed the Sparse Fast Hadamard Transform (SparseFHT) algorithm that elegantly computes the $K$-sparse WHT in the absence of noise using ${O}(K\log N)$ samples in time ${O}(K\log^2 N)$. However, the SparseFHT algorithm explicitly exploits the noiseless nature of the problem, and is not equipped to deal with scenarios where the observations are corrupted by noise. Therefore, a question of critical interest is whether this coding-theoretic framework can be made robust to noise. Further, if the answer is yes, what is the extra price that needs to be paid for being robust to noise? In this paper, we show, quite interestingly, that there is {\it no extra price} that needs to be paid for being robust to noise other than a constant factor. In other words, we can maintain the same sample complexity ${O}(K\log N)$ and the computational complexity ${O}(K\log^2 N)$ as those of the noiseless case, using our SParse Robust Iterative Graph-based Hadamard Transform (SPRIGHT) algorithm.

cs.IT↗

arXiv2015

SU(3) Quantum Hall Ferromagnetism in SnTe

The (111) surface of SnTe hosts one isotropic $\barΓ$-centered and three degenerate anisotropic $\bar{M}$- centered Dirac surface states. We predict that a nematic phase with spontaneously broken $C_3$ symmetry will occur in the presence of an external magnetic field when the $N=0$ $\bar{M}$ Landau levels are $1/3$ or $2/3$ filled. The nematic state phase boundary is controlled by a competition between intravalley Coulomb interactions that favor a valley-polarized state, and weaker intervalley scattering processes that increase in relative strength with magnetic field. An in-plane Zeeman field alters the phase diagram by lifting the three-fold $\bar{M}$ Landau level degeneracy, yielding a ground state energy with $2π/3$ periodicity as a function of Zeeman-field orientation angle.

cond-mat.mes-hall↗

arXiv2015

Giant and tunable valley degeneracy splitting in MoTe2

Monolayer transition-metal dichalcogenides possess a pair of degenerate helical valleys in the band structure that exhibit fascinating optical valley polarization. Optical valley polarization, however, is limited by carrier lifetimes of these materials. Lifting the valley degeneracy is therefore an attractive route for achieving valley polarization. It is very challenging to achieve appreciable valley degeneracy splitting with applied magnetic field. We propose a strategy to create giant splitting of the valley degeneracy by proximity-induced Zeeman effect. As a demonstration, our first principles calculations of monolayer MoTe$_2$ on a EuO substrate show that valley splitting over 300 meV can be generated. The proximity coupling also makes interband transition energies valley dependent, enabling valley selection by optical frequency tuning in addition to circular polarization. The valley splitting in the heterostructure is also continuously tunable by rotating substrate magnetization. The giant and tunable valley splitting adds a readily accessible dimension to the valley-spin physics with rich and interesting experimental consequences, and offers a practical avenue for exploring device paradigms based on the intrinsic degrees of freedom of electrons.

cond-mat.mtrl-sci↗

arXiv2015

Impact of Metro Cell Antenna Pattern and Downtilt in Heterogeneous Networks

This article discusses the positive impact metro cell antennas with narrow vertical beamwidth and electrical downtilt can have on heterogeneous cellular networks. Using a model of random cell placement based on Poisson distribution, along with an innovative 3D building model that quantifies blockage due to shadowing, it is demonstrated that network spectral efficiency and average user throughput both increase as vertical beamwidth is decreased and downtilt is applied to metro cell transmission. Moreover, the network becomes more energy efficient. Importantly, these additional gains in network performance can be achieved without any cooperation or exchange of information between macro cell base stations and metro cells.

cs.IT↗

arXiv2015

Exciton-polaritons in transition-metal dichalcogenides and their direct excitation via energy transfer

Excitons, composite electron-hole quasiparticles, are known to play an important role in optoelectronic phenomena in many semiconducting materials. Recent experiments and theory indicate that the band-gap optics of the newly discovered monolayer transition-metal dichalcogenides (TMDs) is dominated by tightly bound valley excitons. The strong interaction of excitons with long-range electromagnetic fields in these 2D systems can significantly affect their intrinsic properties. Here, we develop a semi-classical framework for intrinsic exciton-polaritons in monolayer TMDs that treats their dispersion and radiative decay on the same footing and can incorporate effects of the dielectric environment. It is demonstrated how both inter- and intra-valley long-range interactions influence the dispersion and decay of the polaritonic eigenstates. We also show that exciton-polaritons can be efficiently excited via resonance energy transfer from quantum emitters such as quantum dots, which may be useful for various applications.

cond-mat.mes-hall↗

arXiv2014

Valley Contrasting Magnetoluminescence in Monolayer MoS$_{2}$ Quantum Hall Systems

The valley dependent optical selection rules in recently discovered monolayer group-VI transition metal dichalcogenides (TMDs) make possible optical control of valley polarization, a crucial step towards valleytronic applications. However, in presence of Landaul level(LL) quantization such selection rules are taken over by selection rules between the LLs, which are not necessarily valley contrasting. Using MoS$_{2}$ as an example we show that the spatial inversion-symmetry breaking results in unusual valley dependent inter-LL selection rules, which directly locks polarization to valley. We find a systematic valley splitting for all Landau levels (LLs) in the quantum Hall regime, whose magnitude is linearly proportional to the magnetic field and in comparable with the LL spacing. Consequently, unique plateau structures are found in the optical Hall conductivity, which can be measured by the magneto-optical Faraday rotations.

cond-mat.mes-hall↗

arXiv2014

Spontaneous Layer-Pseudospin Domain Walls in Bilayer Graphene

Bilayer graphene is susceptible to a family of unusual broken symmetry states with spin and valley dependent layer polarization. We report on a microscopic study of the domain walls in these systems, demonstrating that they have interesting microscopic structures related to order-induced topological characters. We use our results to estimate Ginzburg-Landau model parameters and transition temperatures for the ordered states of bilayer graphene.

cond-mat.mes-hall↗

arXiv2014

The edge engineering of topological Bi(111) bilayer

A topological insulator is a novel quantum state, characterized by symmetry-protected non-trivial edge/surface states. Our first-principle simulations show the significant effects of the chemical decoration on edge states of topological Bi(111) bilayer nanoribbon, which remove the trivial edge state and recover the Dirac linear dispersion of topological edge state. By comparing the edge states with and without chemical decoration, the Bi(111) bilayer nanoribbon offers a simple system for assessing conductance fluctuation of edge states. The chemical decoration can also modify the penetration depth and the spin texture of edge states. A low-energy effective model is proposed to explain the distinctive spin texture of Bi(111) bilayer nanoribbon, which breaks the spin-momentum orthogonality along the armchair edge.

cond-mat.mtrl-sci↗

arXiv2013

Valley Carrier Dynamics in Monolayer Molybdenum Disulphide from Helicity Resolved Ultrafast Pump-probe Spectroscopy

We investigate the valley related carrier dynamics in monolayer MoS2 using helicity resolved non-degenerate ultrafast pump-probe spectroscopy at the vicinity of the high-symmetry K point under the temperature down to 78 K. Monolayer MoS2 shows remarkable transient reflection signals, in stark contrast to bilayer and bulk MoS2 due to the enhancement of many-body effect at reduced dimensionality. The helicity resolved ultrafast time-resolved result shows that the valley polarization is preserved for only several ps before scattering process makes it undistinguishable. We suggest that the dynamical degradation of valley polarization is attributable primarily to the exciton trapping by defect states in the exfoliated MoS2 samples. Our experiment and a tight-binding model analysis also show that the perfect valley CD selectivity is fairly robust against disorder at the K point, but quickly decays from the high-symmetry point in the momentum space in the presence of disorder.

physics.optics↗

arXiv2013

Designer Topological Insulators in Superlattices

Gapless Dirac surface states are protected at the interface of topological and normal band insulators. In a binary superlattice bearing such interfaces, we establish that valley-dependent dimerization of symmetry-unrelated Dirac surface states can be exploited to induce topological quantum phase transitions. This mechanism leads to a rich phase diagram that allows us to design strong, weak, and crystalline topological insulators. Our ab initio simulations further demonstrate this mechanism in [111] and [110] superlattices of calcium and tin tellurides.

cond-mat.mtrl-sci↗

arXiv2013

Emergent topological and half semimetallic Dirac Fermions at oxide interfaces

It is highly desirable to combine recent advances in the topological quantum phases with technologically relevant materials. Chromium dioxide (CrO2) is a half-metallic material, widely used in high-end data storage applications. Using first principles calculations, we show via interfacial orbital design that a novel class of half semi-metallic Dirac electronic phase emerges at the interface CrO2 with TiO2 in both thin film and superlattice configurations, with four spin-polarized Dirac points in momentum-space (k-space) band structure. When the spin and orbital degrees of freedom are allowed to couple, the CrO2/TiO2 superlattice becomes a Chern insulator without external fields or additional doping. With topological gaps equivalent to 43 Kelvin and a Chern number 2, the ensuing quantization of Hall conductance to 2e^2/h will enable potential development of these highly industrialized oxides for applications in topologically high fidelity data storage and energy-efficient electronic and spintronic devices.

cond-mat.mtrl-sci↗

arXiv2013

Zeeman Field-Tuned Transitions for Surface Chern Insulators

Mirror symmetric surfaces of a topological crystalline insulator host even number of Dirac surface states. A surface Zeeman field generically gaps these states leading to a quantized anomalous Hall effect. Varying the direction of Zeeman field induces transitions between different surface insulating states with any two Chern numbers between -4 and 4. In the crystal frame the phase boundaries occur for field orientations which are great circles with (111)-like normals on a sphere.

cond-mat.mes-hall↗

arXiv2013

A Framework for Phasor Measurement Placement in Hybrid State Estimation via Gauss-Newton

In this paper, we study the placement of Phasor Measurement Units (PMU) for enhancing hybrid state estimation via the traditional Gauss-Newton method, which uses measurements from both PMU devices and Supervisory Control and Data Acquisition (SCADA) systems. To compare the impact of PMU placements, we introduce a useful metric which accounts for three important requirements in power system state estimation: {\it convergence}, {\it observability} and {\it performance} (COP). Our COP metric can be used to evaluate the estimation performance and numerical stability of the state estimator, which is later used to optimize the PMU locations. In particular, we cast the optimal placement problem in a unified formulation as a semi-definite program (SDP) with integer variables and constraints that guarantee observability in case of measurements loss. Last but not least, we propose a relaxation scheme of the original integer-constrained SDP with randomization techniques, which closely approximates the optimum deployment. Simulations of the IEEE-30 and 118 systems corroborate our analysis, showing that the proposed scheme improves the convergence of the state estimator, while maintaining optimal asymptotic performance.

math.OC↗

arXiv2013

Magnetic control of the valley degree of freedom of massive Dirac fermions with application to transition metal dichalcogenides

We study the valley-dependent magnetic and transport properties of massive Dirac fermions in multivalley systems such as the transition metal dichalcogenides. The asymmetry of the zeroth Landau level between valleys and the enhanced magnetic susceptibility can be attributed to the different orbital magnetic moment tied with each valley. This allows the valley polarization to be controlled by tuning the external magnetic field and the doping level. As a result of this magnetic field induced valley polarization, there exists an extra contribution to the ordinary Hall effect. All these effects can be captured by a low energy effective theory with a valley-orbit coupling term.

cond-mat.mes-hall↗

arXiv2013

Robust Decentralized State Estimation and Tracking for Power Systems via Network Gossiping

This paper proposes a fully decentralized adaptive re-weighted state estimation (DARSE) scheme for power systems via network gossiping. The enabling technique is the proposed Gossip-based Gauss-Newton (GGN) algorithm, which allows to harness the computation capability of each area (i.e. a database server that accrues data from local sensors) to collaboratively solve for an accurate global state. The DARSE scheme mitigates the influence of bad data by updating their error variances online and re-weighting their contributions adaptively for state estimation. Thus, the global state can be estimated and tracked robustly using near-neighbor communications in each area. Compared to other distributed state estimation techniques, our communication model is flexible with respect to reconfigurations and resilient to random failures as long as the communication network is connected. Furthermore, we prove that the Jacobian of the power flow equations satisfies the Lipschitz condition that is essential for the GGN algorithm to converge to the desired solution. Simulations of the IEEE-118 system show that the DARSE scheme can estimate and track online the global power system state accurately, and degrades gracefully when there are random failures and bad data.

cs.DC↗