arXiv subjects

Wei Xu

Publications and source records attributed to Wei Xu.

At least 541 records · Page 30Linked to original sources

arXiv2019

Deep Learning Based Joint Resource Scheduling Algorithms for Hybrid MEC Networks

In this paper, we consider a hybrid mobile edge computing (H-MEC) platform, which includes ground stations (GSs), ground vehicles (GVs) and unmanned aerial vehicle (UAVs), all with mobile edge cloud installed to enable user equipments (UEs) or Internet of thing (IoT) devices with intensive computing tasks to offload. Our objective is to obtain an online offloading algorithm to minimize the energy consumption of all the UEs, by jointly optimizing the positions of GVs and UAVs, user association and resource allocation in real-time, while considering the dynamic environment. To this end, we propose a hybrid deep learning based online offloading (H2O) framework where a large-scale path-loss fuzzy c-means (LSFCM) algorithm is first proposed and used to predict the optimal positions of GVs and UAVs. Secondly, a fuzzy membership matrix U-based particle swarm optimization (U-PSO) algorithm is applied to solve the mixed integer nonlinear programming (MINLP) problems and generate the sample datasets for the deep neural network (DNN) where the fuzzy membership matrix can capture the small-scale fading effects and the information of mutual interference. Thirdly, a DNN with the scheduling layer is introduced to provide user association and computing resource allocation under the practical latency requirement of the tasks and limited available computing resource of H-MEC. In addition, different from traditional DNN predictor, we only input one UE information to the DNN at one time, which will be suitable for the scenarios where the number of UE is varying and avoid the curse of dimensionality in DNN.

eess.SP↗

arXiv2019

Secure Communication for Spatially Sparse Millimeter-Wave Massive MIMO Channels via Hybrid Precoding

In this paper, we investigate secure communication over sparse millimeter-wave (mm-Wave) massive multiple-input multiple-output (MIMO) channels by exploiting the spatial sparsity of legitimate user's channel. We propose a secure communication scheme in which information data is precoded onto dominant angle components of the sparse channel through a limited number of radio-frequency (RF) chains, while artificial noise (AN) is broadcast over the remaining nondominant angles interfering only with the eavesdropper with a high probability. It is shown that the channel sparsity plays a fundamental role analogous to secret keys in achieving secure communication. Hence, by defining two statistical measures of the channel sparsity, we analytically characterize its impact on secrecy rate. In particular, a substantial improvement on secrecy rate can be obtained by the proposed scheme due to the uncertainty, i.e., "entropy", introduced by the channel sparsity which is unknown to the eavesdropper. It is revealed that sparsity in the power domain can always contribute to the secrecy rate. In contrast, in the angle domain, there exists an optimal level of sparsity that maximizes the secrecy rate. The effectiveness of the proposed scheme and derived results are verified by numerical simulations.

cs.IT↗

arXiv2019

RSA: Byzantine-Robust Stochastic Aggregation Methods for Distributed Learning from Heterogeneous Datasets

In this paper, we propose a class of robust stochastic subgradient methods for distributed learning from heterogeneous datasets at presence of an unknown number of Byzantine workers. The Byzantine workers, during the learning process, may send arbitrary incorrect messages to the master due to data corruptions, communication failures or malicious attacks, and consequently bias the learned model. The key to the proposed methods is a regularization term incorporated with the objective function so as to robustify the learning task and mitigate the negative effects of Byzantine attacks. The resultant subgradient-based algorithms are termed Byzantine-Robust Stochastic Aggregation methods, justifying our acronym RSA used henceforth. In contrast to most of the existing algorithms, RSA does not rely on the assumption that the data are independent and identically distributed (i.i.d.) on the workers, and hence fits for a wider class of applications. Theoretically, we show that: i) RSA converges to a near-optimal solution with the learning error dependent on the number of Byzantine workers; ii) the convergence rate of RSA under Byzantine attacks is the same as that of the stochastic gradient descent method, which is free of Byzantine attacks. Numerically, experiments on real dataset corroborate the competitive performance of RSA and a complexity reduction compared to the state-of-the-art alternatives.

cs.LG↗

arXiv2019

Resource Allocation Using Gradient Boosting Aided Deep Q-Network for IoT in C-RANs

In this paper, we investigate dynamic resource allocation (DRA) problems for Internet of Things (IoT) in real-time cloud radio access networks (C-RANs), by combining gradient boosting approximation and deep reinforcement learning to solve the following two major problems. Firstly, in C-RANs, the decision making process of resource allocation is time-consuming and computational-expensive, motivating us to use an approximation method, i.e. the gradient boosting decision tree (GBDT) to approximate the solutions of second order cone programming (SOCP) problem. Moreover, considering the innumerable states in real-time C-RAN systems, we employ a deep reinforcement learning framework, i.e., deep Q-network (DQN) to generate a robust policy that controls the status of remote radio heads (RRHs). We propose a GBDT-based DQN framework for the DRA problem, where the heavy computation to solve SOCP problems is cut down and great power consumption is saved in the whole C-RAN system. We demonstrate that the generated policy is error-tolerant even the gradient boosting regression may not be strictly subject to the constraints of the original problem. Comparisons between the proposed method and existing baseline methods confirm the advantages of our method.

cs.IT↗

arXiv2019

Visual Understanding of Multiple Attributes Learning Model of X-Ray Scattering Images

This extended abstract presents a visualization system, which is designed for domain scientists to visually understand their deep learning model of extracting multiple attributes in x-ray scattering images. The system focuses on studying the model behaviors related to multiple structural attributes. It allows users to explore the images in the feature space, the classification output of different attributes, with respect to the actual attributes labelled by domain scientists. Abundant interactions allow users to flexibly select instance images, their clusters, and compare them visually in details. Two preliminary case studies demonstrate its functionalities and usefulness.

cs.LG↗

arXiv2019

Identifying Candidate Spaces for Advert Implantation

Virtual advertising is an important and promising feature in the area of online advertising. It involves integrating adverts onto live or recorded videos for product placements and targeted advertisements. Such integration of adverts is primarily done by video editors in the post-production stage, which is cumbersome and time-consuming. Therefore, it is important to automatically identify candidate spaces in a video frame, wherein new adverts can be implanted. The candidate space should match the scene perspective, and also have a high quality of experience according to human subjective judgment. In this paper, we propose the use of a bespoke neural net that can assist the video editors in identifying candidate spaces. We benchmark our approach against several deep-learning architectures on a large-scale image dataset of candidate spaces of outdoor scenes. Our work is the first of its kind in this area of multimedia and augmented reality applications, and achieves the best results.

cs.CV↗

arXiv2019

Bit-level Optimized Neural Network for Multi-antenna Channel Quantization

Quantized channel state information (CSI) plays a critical role in precoding design which helps reap the merits of multiple-input multiple-output (MIMO) technology. In order to reduce the overhead of CSI feedback, we propose a deep learning based CSI quantization method by developing a joint convolutional residual network (JCResNet) which benefits MIMO channel feature extraction and recovery from the perspective of bit-level quantization performance. Experiments show that our proposed method substantially improves the performance.

cs.IT↗

arXiv2019

Doc2EDAG: An End-to-End Document-level Framework for Chinese Financial Event Extraction

Most existing event extraction (EE) methods merely extract event arguments within the sentence scope. However, such sentence-level EE methods struggle to handle soaring amounts of documents from emerging applications, such as finance, legislation, health, etc., where event arguments always scatter across different sentences, and even multiple such event mentions frequently co-exist in the same document. To address these challenges, we propose a novel end-to-end model, Doc2EDAG, which can generate an entity-based directed acyclic graph to fulfill the document-level EE (DEE) effectively. Moreover, we reformalize a DEE task with the no-trigger-words design to ease the document-level event labeling. To demonstrate the effectiveness of Doc2EDAG, we build a large-scale real-world dataset consisting of Chinese financial announcements with the challenges mentioned above. Extensive experiments with comprehensive analyses illustrate the superiority of Doc2EDAG over state-of-the-art methods. Data and codes can be found at https://github.com/dolphin-zs/Doc2EDAG.

cs.CL↗

arXiv2019

CAMEL: A Weakly Supervised Learning Framework for Histopathology Image Segmentation

Histopathology image analysis plays a critical role in cancer diagnosis and treatment. To automatically segment the cancerous regions, fully supervised segmentation algorithms require labor-intensive and time-consuming labeling at the pixel level. In this research, we propose CAMEL, a weakly supervised learning framework for histopathology image segmentation using only image-level labels. Using multiple instance learning (MIL)-based label enrichment, CAMEL splits the image into latticed instances and automatically generates instance-level labels. After label enrichment, the instance-level labels are further assigned to the corresponding pixels, producing the approximate pixel-level labels and making fully supervised training of segmentation models possible. CAMEL achieves comparable performance with the fully supervised approaches in both instance-level classification and pixel-level segmentation on CAMELYON16 and a colorectal adenoma dataset. Moreover, the generality of the automatic labeling methodology may benefit future weakly supervised learning studies for histopathology image analysis.

eess.IV↗

arXiv2019

Functional Limit Theorems for Marked Hawkes Point Measures

This paper establishes a functional law of large numbers and a functional central limit theorem for marked Hawkes point measures and their corresponding shot noise processes. We prove that the normalized random measure can be approximated in distribution by the sum of a Gaussian white noise process plus an appropriate lifting map of a correlated one-dimensional Brownian motion. The Brownian results from the self-exiting arrivals of events. We apply our limit theorems for Hawkes point measures to analyze the population dynamics of budding microbes in a host.

math.PR↗

arXiv2019

Ergodic Rate Analysis of Cooperative Ambient Backscatter Communication

Ambient backscatter communication has shown great potential in the development of future wireless networks. It enables a backscatter transmitter (BTx) to send information directly to an adjacent receiver by modulating over ambient radio frequency (RF) carriers. In this paper, we consider a cooperative ambient backscatter communication system where a multi-antenna cooperative receiver separately decodes signals from an RF source and a BTx. Upper bounds of the ergodic rates of both links are derived. The power scaling laws are accordingly characterized for both the primary cellular transmission and the cooperative backscatter. The impact of additional backscatter link is also quantitatively analyzed. Simulation results are provided to verify the derived results.

cs.IT↗

arXiv2019

Weighted Spectral Efficiency Optimization for Hybrid Beamforming in Multiuser Massive MIMO-OFDM Systems

In this paper, we consider hybrid beamforming designs for multiuser massive multiple-input multiple-output (MIMO)-orthogonal frequency division multiplexing (OFDM) systems. Aiming at maximizing the weighted spectral efficiency, we propose one alternating maximization framework where the analog precoding is optimized by Riemannian manifold optimization. If the digital precoding is optimized by a locally optimal algorithm, we obtain a locally optimal alternating maximization algorithm. In contrast, if we use a weighted minimum mean square error (MMSE)-based iterative algorithm for digital precoding, we obtain a suboptimal alternating maximization algorithm with reduced complexity in each iteration. By characterizing the upper bound of the weighted arithmetic and geometric means of mean square errors (MSEs), it is shown that the two alternating maximization algorithms have similar performance when the user specific weights do not have big differences. Verified by numerical results, the performance gap between the two alternating maximization algorithms becomes large when the ratio of the maximal and minimal weights among users is very large. Moreover, we also propose a low-complexity closed-form method without iterations. It employs matrix decomposition for the analog beamforming and weighted MMSE for the digital beamforming. Although it is not supposed to maximize the weighted spectral efficiency, it exhibits small performance deterioration compared to the two iterative alternating maximization algorithms and it qualifies as a good initialization for iterative algorithms, saving thereby iterations.

eess.SP↗

arXiv2019

Cache Placement in Two-Tier HetNets with Limited Storage Capacity: Cache or Buffer?

In this paper, we aim to minimize the average file transmission delay via bandwidth allocation and cache placement in two-tier heterogeneous networks with limited storage capacity, which consists of cache capacity and buffer capacity. For average delay minimization problem with fixed bandwidth allocation, although this problem is nonconvex, the optimal solution is obtained in closed form by comparing all locally optimal solutions calculated from solving the Karush-Kuhn-Tucker conditions. To jointly optimize bandwidth allocation and cache placement, the optimal bandwidth allocation is first derived and then substituted into the original problem. The structure of the optimal caching strategy is presented, which shows that it is optimal to cache the files with high popularity instead of the files with big size. Based on this optimal structure, we propose an iterative algorithm with low complexity to obtain a suboptimal solution, where the closed-from expression is obtained in each step. Numerical results show the superiority of our solution compared to the conventional cache strategy without considering cache and buffer tradeoff in terms of delay.

cs.IT↗

arXiv2019

Every Pixel Counts ++: Joint Learning of Geometry and Motion with 3D Holistic Understanding

Learning to estimate 3D geometry in a single frame and optical flow from consecutive frames by watching unlabeled videos via deep convolutional network has made significant progress recently. Current state-of-the-art (SoTA) methods treat the two tasks independently. One typical assumption of the existing depth estimation methods is that the scenes contain no independent moving objects. while object moving could be easily modeled using optical flow. In this paper, we propose to address the two tasks as a whole, i.e. to jointly understand per-pixel 3D geometry and motion. This eliminates the need of static scene assumption and enforces the inherent geometrical consistency during the learning process, yielding significantly improved results for both tasks. We call our method as "Every Pixel Counts++" or "EPC++". Specifically, during training, given two consecutive frames from a video, we adopt three parallel networks to predict the camera motion (MotionNet), dense depth map (DepthNet), and per-pixel optical flow between two frames (OptFlowNet) respectively. The three types of information are fed into a holistic 3D motion parser (HMP), and per-pixel 3D motion of both rigid background and moving objects are disentangled and recovered. Comprehensive experiments were conducted on datasets with different scenes, including driving scenario (KITTI 2012 and KITTI 2015 datasets), mixed outdoor/indoor scenes (Make3D) and synthetic animation (MPI Sintel dataset). Performance on the five tasks of depth estimation, optical flow estimation, odometry, moving object segmentation and scene flow estimation shows that our approach outperforms other SoTA methods. Code will be available at: https://github.com/chenxuluo/EPC.

cs.CV↗

arXiv2019

Highly degenerate photonic flat bands arising from complete graph configurations

Inspired by complete graph theory, we demonstrate that a metallic claw "meta-atom" structure can carry a high number of nearly degenerate resonant modes. A photonic meta-crystal composing of a lattice of such meta-atoms exhibits a large number of flat bands that are squeezed into a narrow frequency window, and these flat bands can be designed to locate in a wide complete 3D bandgap. The degeneracy dimension (Nf) of the flat bands is determined by the number of branches (Nb) of the metallic claw with Nf=Nb-3, which is geometrically related to the complete graph theory. Different from those flat bands emerging from special lattice arrangements (e.g., Kagome lattice), the isolated flat bands here are insensitive to lattice perturbations. The proposed mechanism offers a new platform for realizing various dispersion-less phenomena and a new paradigm to realize high density of states and spectra compressing.

physics.class-ph↗

arXiv2019

Multi-task Pairwise Neural Ranking for Hashtag Segmentation

Hashtags are often employed on social media and beyond to add metadata to a textual utterance with the goal of increasing discoverability, aiding search, or providing additional semantics. However, the semantic content of hashtags is not straightforward to infer as these represent ad-hoc conventions which frequently include multiple words joined together and can include abbreviations and unorthodox spellings. We build a dataset of 12,594 hashtags split into individual segments and propose a set of approaches for hashtag segmentation by framing it as a pairwise ranking problem between candidate segmentations. Our novel neural approaches demonstrate 24.6% error reduction in hashtag segmentation accuracy compared to the current state-of-the-art method. Finally, we demonstrate that a deeper understanding of hashtag semantics obtained through segmentation is useful for downstream applications such as sentiment analysis, for which we achieved a 2.6% increase in average recall on the SemEval 2017 sentiment analysis dataset.

cs.CL↗

arXiv2019

A new data fitting method for stretched Gaussian noise: stretched least square method

Stretched Gaussian distribution is the fundamental solution of the Hausdorff derivative diffusion equation and its corresponding stretched Gaussian noise is a widely encountered non-Gaussian noise in science and engineering. The least square method is a standard regression approach to fit Gaussian noisy data, but has distinct limits for non-Gaussian noise. Based on the Hausdorff calculus, this study develops a stretched least square method to fit stretched Gaussian noise by using the Hausdorff fractal distance as the horizontal coordinate. To better compare with the least square method, different high levels of stretched Gaussian noise are added to real values. Numerical experiments show the stretched least square method is more accurate specific in fitting stretched Gaussian noise than the least square method.

math.NA↗

arXiv2019

DIAG-NRE: A Neural Pattern Diagnosis Framework for Distantly Supervised Neural Relation Extraction

Pattern-based labeling methods have achieved promising results in alleviating the inevitable labeling noises of distantly supervised neural relation extraction. However, these methods require significant expert labor to write relation-specific patterns, which makes them too sophisticated to generalize quickly.To ease the labor-intensive workload of pattern writing and enable the quick generalization to new relation types, we propose a neural pattern diagnosis framework, DIAG-NRE, that can automatically summarize and refine high-quality relational patterns from noise data with human experts in the loop. To demonstrate the effectiveness of DIAG-NRE, we apply it to two real-world datasets and present both significant and interpretable improvements over state-of-the-art methods.

cs.CL↗