arXiv subjects

Ang Li

Publications and source records attributed to Ang Li.

At least 559 records · Page 31Linked to original sources

arXiv2019

Hard superconducting gap and diffusion-induced superconductors in Ge-Si nanowires

We show a hard induced superconducting gap in a Ge-Si nanowire Josephson transistor up to in-plane magnetic fields of $250$ mT, an important step towards creating and detecting Majorana zero modes in this system. A hard induced gap requires a highly homogeneous tunneling heterointerface between the superconducting contacts and the semiconducting nanowire. This is realized by annealing devices at $180$ $^\circ$C during which aluminium inter-diffuses and replaces the germanium in a section of the nanowire. Next to Al, we find a superconductor with lower critical temperature ($T_\mathrm{C}=0.9$ K) and a higher critical field ($B_\mathrm{C}=0.9-1.2$ T). We can therefore selectively switch either superconductor to the normal state by tuning the temperature and the magnetic field and observe that the additional superconductor induces a proximity supercurrent in the semiconducting part of the nanowire even when the Al is in the normal state. In another device where the diffusion of Al rendered the nanowire completely metallic, a superconductor with a much higher critical temperature ($T_\mathrm{C}=2.9$ K) and critical field ($B_\mathrm{C}=3.4$ T) is found. The small size of diffusion-induced superconductors inside nanowires may be of special interest for applications requiring high magnetic fields in arbitrary direction.

cond-mat.mes-hall↗

arXiv2019

Improved Knowledge Distillation via Teacher Assistant

Despite the fact that deep neural networks are powerful models and achieve appealing results on many tasks, they are too large to be deployed on edge devices like smartphones or embedded sensor nodes. There have been efforts to compress these networks, and a popular method is knowledge distillation, where a large (teacher) pre-trained network is used to train a smaller (student) network. However, in this paper, we show that the student network performance degrades when the gap between student and teacher is large. Given a fixed student network, one cannot employ an arbitrarily large teacher, or in other words, a teacher can effectively transfer its knowledge to students up to a certain size, not smaller. To alleviate this shortcoming, we introduce multi-step knowledge distillation, which employs an intermediate-sized network (teacher assistant) to bridge the gap between the student and the teacher. Moreover, we study the effect of teacher assistant size and extend the framework to multi-step distillation. Theoretical analysis and extensive experiments on CIFAR-10,100 and ImageNet datasets and on CNN and ResNet architectures substantiate the effectiveness of our proposed approach.

cs.LG↗

arXiv2019

Hybrid Precoding Design for Reconfigurable Intelligent Surface aided mmWave Communication Systems

In this letter, we focus on the hybrid precoding (HP) design for the reconfigurable intelligent surface (RIS) aided multi-user (MU) millimeter wave (mmWave) communication systems. Specifically, we aim to minimize the mean-squared-error (MSE) between the received symbols and the transmitted symbols by jointly optimizing the analog-digital HP at the base-station (BS) and the phase shifts (PSs) at the RIS, where the non-convex element-wise constant-modulus constraints for the analog precoding and the PSs are tackled by resorting to the gradient-projection (GP) method. We analytically prove the convergence of the proposed algorithm and demonstrate the desirable performance gain for the proposed design through numerical results.

cs.NI↗

arXiv2019

Anisotropies of the g-factor tensor and diamagnetic coefficient in crystal-phase quantum dots in InP nanowires

Crystal-phase low-dimensional structures offer great potential for the implementation of photonic devices of interest for quantum information processing. In this context, unveiling the fundamental parameters of the crystal phase structure is of much relevance for several applications. Here, we report on the anisotropy of the g-factor tensor and diamagnetic coefficient in wurtzite/zincblende (WZ/ZB) crystal-phase quantum dots (QDs) realized in single InP nanowires. The WZ and ZB alternating axial sections in the NWs are identified by high-angle annular dark-field scanning transmission electron microscopy. The electron (hole) g-factor tensor and the exciton diamagnetic coefficients in WZ/ZB crystal-phase QDs are determined through micro-photoluminescence measurements at low temperature (4.2 K) with different magnetic field configurations, and rationalized by invoking the spin-correlated orbital current model. Our work provides key parameters for band gap engineering and spin states control in crystal-phase low-dimensional structures in nanowires.

cond-mat.mes-hall↗

arXiv2019

Orthogonal Gradient Descent for Continual Learning

Neural networks are achieving state of the art and sometimes super-human performance on learning tasks across a variety of domains. Whenever these problems require learning in a continual or sequential manner, however, neural networks suffer from the problem of catastrophic forgetting; they forget how to solve previous tasks after being trained on a new task, despite having the essential capacity to solve both tasks if they were trained on both simultaneously. In this paper, we propose to address this issue from a parameter space perspective and study an approach to restrict the direction of the gradient updates to avoid forgetting previously-learned data. We present the Orthogonal Gradient Descent (OGD) method, which accomplishes this goal by projecting the gradients from new tasks onto a subspace in which the neural network output on previous task does not change and the projected gradient is still in a useful direction for learning the new task. Our approach utilizes the high capacity of a neural network more efficiently and does not require storing the previously learned data that might raise privacy concerns. Experiments on common benchmarks reveal the effectiveness of the proposed OGD method.

cs.LG↗

arXiv2019

Computation Offloading for IoT in C-RAN: Optimization and Deep Learning

We consider computation offloading for Internet-of-things (IoT) applications in multiple-input-multiple-output (MIMO) cloud-radio-access-network (C-RAN). Due to the limited battery life and computational capability in the IoT devices (IoTDs), the computational tasks of the IoTDs are offloaded to a MIMO C-RAN, where a MIMO radio resource head (RRH) is connected to a baseband unit (BBU) through a capacity-limited fronthaul link, facilitated by the spatial filtering and uniform scalar quantization. We formulate a computation offloading optimization problem to minimize the total transmit power of the IoTDs while satisfying the latency requirement of the computational tasks, and find that the problem is non-convex. To obtain a feasible solution, firstly the spatial filtering matrix is locally optimized at the MIMO RRH. Subsequently, we leverage the alternating optimization framework for joint optimization on the residual variables at the BBU, where the baseband combiner is obtained in a closed-form, the resource allocation sub-problem is solved through successive inner convexification, and the number of quantization bits is obtained by a line-search method. As a low-complexity approach, we deploy a supervised deep learning method, which is trained with the solutions to our optimization algorithm. Numerical results validate the effectiveness of the proposed algorithm and the deep learning method.

cs.NI↗

arXiv2019

Cross-View Policy Learning for Street Navigation

The ability to navigate from visual observations in unfamiliar environments is a core component of intelligent agents and an ongoing challenge for Deep Reinforcement Learning (RL). Street View can be a sensible testbed for such RL agents, because it provides real-world photographic imagery at ground level, with diverse street appearances; it has been made into an interactive environment called StreetLearn and used for research on navigation. However, goal-driven street navigation agents have not so far been able to transfer to unseen areas without extensive retraining, and relying on simulation is not a scalable solution. Since aerial images are easily and globally accessible, we propose instead to train a multi-modal policy on ground and aerial views, then transfer the ground view policy to unseen (target) parts of the city by utilizing aerial view observations. Our core idea is to pair the ground view with an aerial view and to learn a joint policy that is transferable across views. We achieve this by learning a similar embedding space for both views, distilling the policy across views and dropping out visual modalities. We further reformulate the transfer learning paradigm into three stages: 1) cross-modal training, when the agent is initially trained on multiple city regions, 2) aerial view-only adaptation to a new area, when the agent is adapted to a held-out region using only the easily obtainable aerial view, and 3) ground view-only transfer, when the agent is tested on navigation tasks on unseen ground views, without aerial imagery. Experimental results suggest that the proposed cross-view policy learning enables better generalization of the agent and allows for more effective transfer to unseen environments.

cs.CV↗

arXiv2019

Secure Interference Exploitation Precoding in MISO Wiretap Channel: Destructive Region Redefinition with Efficient Solutions

In this paper, we focus on the physical layer security for a K-user multiple-input-single-output (MISO) wiretap channel in the presence of a malicious eavesdropper, where we propose several interference exploitation (IE) precoding schemes for different types of the eavesdropper. Specifically, in the case where a common eavesdropper decodes the signal directly and Eve's full channel state information (CSI) is available at the transmitter, we show that the required transmit power can be further reduced by re-designing the `destructive region' of the constellations for symbol-level precoding and re-formulating the power minimization problem. We further study the SINR balancing problems with the derived `complete destructive region' with full, statistical and no Eve's CSI, respectively, and show that the SINR balancing problem becomes non-convex with statistical or no Eve's CSI. On the other hand, in the presence of a smart eavesdropper using maximal likelihood (ML) detection, the security cannot be guaranteed with all the existing approaches. To this end, we further propose a random jamming scheme (RJS) and a random precoding scheme (RPS), respectively. To solve the introduced convex/non-convex problems in an efficient manner, we propose an iterative algorithm for the convex ones based on the Karush-Kuhn-Tucker (KKT) conditions, and deal with the non-convex ones by resorting to Taylor expansions. Simulation results show that all proposed schemes outperform the existing works in secrecy performance, and that the proposed algorithm improves the computation efficiency significantly.

eess.SP↗

arXiv2019

Flat band magnetism and helical magnetic order in Ni-doped SrCo$_2$As$_2$

A series of Sr(Co$_{1-x}$Ni$_x$)$_2$As$_2$ single crystals was synthesized allowing a comprehensive phase diagram with respect to field, temperature, and chemical substitution to be established. Our neutron diffraction experiments revealed a helimagnetic order with magnetic moments ferromagnetically (FM) aligned in the $ab$ plane and a helimagnetic wavevector of $q=(0,0,0.56)$ for $x$ = 0.1. The combination of neutron diffraction and angle-resolved photoemission spectroscopy (ARPES) measurements show that the tuning of a flat band with $d_{x^2-y^2}$ orbital character drives the helimagnetism and indicates the possibility of a quantum order-by-disorder mechanism.

cond-mat.str-el↗

arXiv2019

Multiple Andreev reflections and Shapiro steps in a Ge-Si nanowire Josephson junction

We present a Josephson junction based on a Ge-Si core-shell nanowire with transparent superconducting Al contacts, a building block which could be of considerable interest for investigating Majorana bound states, superconducting qubits and Andreev (spin) qubits. We demonstrate the dc Josephson effect in the form of a finite supercurrent through the junction, and establish the ac Josephson effect by showing up to 23 Shapiro steps. We observe multiple Andreev reflections up to the sixth order, indicating that charges can scatter elastically many times inside our junction, and that our interfaces between superconductor and semiconductor are transparent and have low disorder.

cond-mat.mes-hall↗

arXiv2019

Boosted GAN with Semantically Interpretable Information for Image Inpainting

Image inpainting aims at restoring missing region of corrupted images, which has many applications such as image restoration and object removal. However, current GAN-based inpainting models fail to explicitly consider the semantic consistency between restored images and original images. Forexample, given a male image with image region of one eye missing, current models may restore it with a female eye. This is due to the ambiguity of GAN-based inpainting models: these models can generate many possible restorations given a missing region. To address this limitation, our key insight is that semantically interpretable information (such as attribute and segmentation information) of input images (with missing regions) can provide essential guidance for the inpainting process. Based on this insight, we propose a boosted GAN with semantically interpretable information for image inpainting that consists of an inpainting network and a discriminative network. The inpainting network utilizes two auxiliary pretrained networks to discover the attribute and segmentation information of input images and incorporates them into the inpainting process to provide explicit semantic-level guidance. The discriminative network adopts a multi-level design that can enforce regularizations not only on overall realness but also on attribute and segmentation consistency with the original images. Experimental results show that our proposed model can preserve consistency on both attribute and segmentation level, and significantly outperforms the state-of-the-art models.

cs.CV↗

arXiv2019

Interference Exploitation 1-Bit Massive MIMO Precoding: A Partial Branch-and-Bound Solution with Near-Optimal Performance

In this paper, we focus on 1-bit precoding approaches for downlink massive multiple-input multiple-output (MIMO) systems, where we exploit the concept of constructive interference (CI). For both PSK and QAM signaling, we firstly formulate the optimization problem that maximizes the CI effect subject to the requirement of the 1-bit transmit signals. We then mathematically prove that, when employing the CI formulation and relaxing the 1-bit constraint, the majority of the transmit signals already satisfy the 1-bit formulation. Building upon this important observation, we propose a 1-bit precoding approach that further improves the performance of the conventional 1-bit CI precoding via a partial branch-and-bound (P-BB) process, where the BB procedure is performed only for the entries that do not comply with the 1-bit requirement. This operation allows a significant complexity reduction compared to the fully-BB (F-BB) process, and enables the BB framework to be applicable to the complex massive MIMO scenarios. We further develop an alternative 1-bit scheme through an `Ordered Partial Sequential Update' (OPSU) process that allows an additional complexity reduction. Numerical results show that both proposed 1-bit precoding methods exhibit a significant signal-to-noise ratio (SNR) gain for the error rate performance, especially for higher-order modulations.

eess.SP↗

arXiv2019

Rethinking Classification and Localization for Cascade R-CNN

We extend the state-of-the-art Cascade R-CNN with a simple feature sharing mechanism. Our approach focuses on the performance increases on high IoU but decreases on low IoU thresholds--a key problem this detector suffers from. Feature sharing is extremely helpful, our results show that given this mechanism embedded into all stages, we can easily narrow the gap between the last stage and preceding stages on low IoU thresholds without resorting to the commonly used testing ensemble but the network itself. We also observe obvious improvements on all IoU thresholds benefited from feature sharing, and the resulting cascade structure can easily match or exceed its counterparts, only with negligible extra parameters introduced. To push the envelope, we demonstrate 43.2 AP on COCO object detection without any bells and whistles including testing ensemble, surpassing previous Cascade R-CNN by a large margin. Our framework is easy to implement and we hope it can serve as a general and strong baseline for future research.

cs.CV↗

arXiv2019

Note on neutron star equation of state in the light of GW170817

From the very first multimessenger event of GW170817, clean robust constraints can be obtained for the tidal deformabilities of the two stars involved in the merger, which provides us unique opportunity to study the equation of states (EOSs) of dense stellar matter. In this contribution, we employ a model from the quark level, describing consistently a nucleon and many-body nucleonic system from a quark potential. We check that our sets of EOSs are consistent with available experimental and observational constraints at both sub-nuclear saturation densities and higher densities. The agreements with ab-initio calculations are also good. Especially, we tune the density dependence of the symmetry energy (characterized by its slope at nuclear saturation $L$) and study its influence on the tidal deformability. The so-called $QMF18$ EOS is named after the case of $L=40~\rm MeV$, and it gives $M_{\rm TOV} =2.08~M_\odot$ and $R= 11.77~\rm km$, $Λ=331$ for a $1.4\,M_\odot$ star. The tidal signals are demonstrated to be insensitive to the uncertain crust-core matching, despite the good correlation between the symmetry energy slope and the radius of the star.

nucl-th↗

arXiv2019

Interference Exploitation via Symbol-Level Precoding: Overview, State-of-the-Art and Future Directions

Interference is traditionally viewed as a performance limiting factor in wireless communication systems, which is to be minimized or mitigated. Nevertheless, a recent line of work has shown that by manipulating the interfering signals such that they add up constructively at the receiver side, known interference can be made beneficial and further improve the system performance in a variety of wireless scenarios, achieved by symbol-level precoding (SLP). This paper aims to provide a tutorial on interference exploitation techniques from the perspective of precoding design in a multi-antenna wireless communication system, by beginning with the classification of constructive interference (CI) and destructive interference (DI). The definition for CI is presented and the corresponding mathematical characterization is formulated for popular modulation types, based on which optimization-based precoding techniques are discussed. In addition, the extension of CI precoding to other application scenarios as well as for hardware efficiency is also described. Proof-of-concept testbeds are demonstrated for the potential practical implementation of CI precoding, and finally a list of open problems and practical challenges are presented to inspire and motivate further research directions in this area.

eess.SP↗

arXiv2019

Beam Allocation for Millimeter-Wave MIMO Tracking Systems

In this paper, we propose a new beam allocation strategy aiming to maximize the average successful tracking probability (ASTP) of time-varying millimeter-wave MIMO systems. In contrast to most existing works that employ one transmitting-receiving (Tx-Rx) beam pair once only in each training period, we investigate a more general framework, where the Tx-Rx beam pairs are allowed to be used repeatedly to improve the received signal powers in specific directions. In the case of orthogonal Tx-Rx beam pairs, a power-based estimator is employed to track the time-varying AoA and AoD of the channel, and the resulting training beam pair sequence design problem is formulated as an integer nonlinear programming (I-NLP) problem. By dividing the feasible region into a set of subregions, the formulated I-NLP is decomposed into a series of concave sub I-NLPs, which can be solved by recursively invoking a nonlinear branch-and-bound algorithm. To reduce the computational cost, we relax the integer constraints of each sub I-NLP and obtain a low-complexity solution via solving the Karush-Kuhn-Tucker conditions of their relaxed problems. For the case when the Tx-Rx beam pairs are overlapped in the angular space, we estimate the updated AoA and AoD via an orthogonal matching pursuit (OMP) algorithm. Moreover, since no explicit expression for the ASTP exists for the OMP-based estimator, we derive a closed-form lower bound of the ASTP, based on which a favorable beam pair allocation strategy can be obtained. Numerical results demonstrate the superiority of the proposed beam allocation strategy over existing benchmarks.

eess.SP↗

arXiv2019

VBSCan Thessaloniki 2018 Workshop Summary

This document reports the first year of activity of the VBSCan COST Action network, as summarised by the talks and discussions happened during the VBSCan Thessaloniki 2018 workshop. The VBSCan COST action is aiming at a consistent and coordinated study of vector-boson scattering from the phenomenological and experimental point of view, for the best exploitation of the data that will be delivered by existing and future particle colliders.

hep-ph↗

arXiv2019

Strain engineering in Ge/GeSn core/shell nanowires

Strain engineering in Sn-rich group IV semiconductors is a key enabling factor to exploit the direct band gap at mid-infrared wavelengths. Here, we investigate the effect of strain on the growth of GeSn alloys in a Ge/GeSn core/shell nanowire geometry. Incorporation of Sn content in the 10-20 at.% range is achieved with Ge core diameters ranging from 50nm to 100nm. While the smaller cores lead to the formation of a regular and homogeneous GeSn shell, larger cores lead to the formation of multi-faceted sidewalls and broadened segregation domains, inducing the nucleation of defects. This behavior is rationalized in terms of the different residual strain, as obtained by realistic finite element method simulations. The extended analysis of the strain relaxation as a function of core and shell sizes, in comparison with the conventional planar geometry, provides a deeper understanding of the role of strain in the epitaxy of metastable GeSn semiconductors.

physics.app-ph↗