arXiv subjects

Pan Zhou

Publications and source records attributed to Pan Zhou.

288 records · Page 16Linked to original sources

arXiv2018

Joint Service Caching and Task Offloading for Mobile Edge Computing in Dense Networks

Mobile Edge Computing (MEC) pushes computing functionalities away from the centralized cloud to the network edge, thereby meeting the latency requirements of many emerging mobile applications and saving backhaul network bandwidth. Although many existing works have studied computation offloading policies, service caching is an equally, if not more important, design topic of MEC, yet receives much less attention. Service caching refers to caching application services and their related databases/libraries in the edge server (e.g. MEC-enabled BS), thereby enabling corresponding computation tasks to be executed. Because only a small number of application services can be cached in resource-limited edge server at the same time, which services to cache has to be judiciously decided to maximize the edge computing performance. In this paper, we investigate the extremely compelling but much less studied problem of dynamic service caching in MEC-enabled dense cellular networks. We propose an efficient online algorithm, called OREO, which jointly optimizes dynamic service caching and task offloading to address a number of key challenges in MEC systems, including service heterogeneity, unknown system dynamics, spatial demand coupling and decentralized coordination. Our algorithm is developed based on Lyapunov optimization and Gibbs sampling, works online without requiring future information, and achieves provable close-to-optimal performance. Simulation results show that our algorithm can effectively reduce computation latency for end users while keeping energy consumption low.

cs.DC↗

arXiv2017

Jointly Attentive Spatial-Temporal Pooling Networks for Video-based Person Re-Identification

Person Re-Identification (person re-id) is a crucial task as its applications in visual surveillance and human-computer interaction. In this work, we present a novel joint Spatial and Temporal Attention Pooling Network (ASTPN) for video-based person re-identification, which enables the feature extractor to be aware of the current input video sequences, in a way that interdependency from the matching items can directly influence the computation of each other's representation. Specifically, the spatial pooling layer is able to select regions from each frame, while the attention temporal pooling performed can select informative frames over the sequence, both pooling guided by the information from distance matching. Experiments are conduced on the iLIDS-VID, PRID-2011 and MARS datasets and the results demonstrate that this approach outperforms existing state-of-art methods. We also analyze how the joint pooling in both dimensions can boost the person re-id performance more effectively than using either of them separately.

cs.CV↗

arXiv2017

Optimal Repair Layering for Erasure-Coded Data Centers: From Theory to Practice

Repair performance in hierarchical data centers is often bottlenecked by cross-rack network transfer. Recent theoretical results show that the cross-rack repair traffic can be minimized through repair layering, whose idea is to partition a repair operation into inner-rack and cross-rack layers. However, how repair layering should be implemented and deployed in practice remains an open issue. In this paper, we address this issue by proposing a practical repair layering framework called DoubleR. We design two families of practical double regenerating codes (DRC), which not only minimize the cross-rack repair traffic, but also have several practical properties that improve state-of-the-art regenerating codes. We implement and deploy DoubleR atop Hadoop Distributed File System (HDFS), and show that DoubleR maintains the theoretical guarantees of DRC and improves the repair performance of regenerating codes in both node recovery and degraded read operations.

cs.DC↗

arXiv2017

The Landscape of Deep Learning Algorithms

This paper studies the landscape of empirical risk of deep neural networks by theoretically analyzing its convergence behavior to the population risk as well as its stationary points and properties. For an $l$-layer linear neural network, we prove its empirical risk uniformly converges to its population risk at the rate of $\mathcal{O}(r^{2l}\sqrt{d\log(l)}/\sqrt{n})$ with training sample size of $n$, the total weight dimension of $d$ and the magnitude bound $r$ of weight of each layer. We then derive the stability and generalization bounds for the empirical risk based on this result. Besides, we establish the uniform convergence of gradient of the empirical risk to its population counterpart. We prove the one-to-one correspondence of the non-degenerate stationary points between the empirical and population risks with convergence guarantees, which describes the landscape of deep neural networks. In addition, we analyze these properties for deep nonlinear neural networks with sigmoid activation functions. We prove similar results for convergence behavior of their empirical risks as well as the gradients and analyze properties of their non-degenerate stationary points. To our best knowledge, this work is the first one theoretically characterizing landscapes of deep learning algorithms. Besides, our results provide the sample complexity of training a good deep neural network. We also provide theoretical understanding on how the neural network depth $l$, the layer width, the network size $d$ and parameter magnitude determine the neural network landscapes.

stat.ML↗

arXiv2016

Private and Truthful Aggregative Game for Large-Scale Spectrum Sharing

Thanks to the rapid development of information technology, the size of the wireless network becomes larger and larger, which makes spectrum resources more precious than ever before. To improve the efficiency of spectrum utilization, game theory has been applied to study the spectrum sharing in wireless networks for a long time. However, the scale of wireless network in existing studies is relatively small. In this paper, we introduce a novel game and model the spectrum sharing problem as an aggregative game for large-scale, heterogeneous, and dynamic networks. The massive usage of spectrum also leads to easier privacy divulgence of spectrum users' actions, which calls for privacy and truthfulness guarantees in wireless network. In a large decentralized scenario, each user has no priori about other users' decisions, which forms an incomplete information game. A "weak mediator", e.g., the base station or licensed spectrum regulator, is introduced and turns this game into a complete one, which is essential to reach a Nash equilibrium (NE). By utilizing past experience on the channel access, we propose an online learning algorithm to improve the utility of each user, achieving NE over time. Our learning algorithm also provides no regret guarantee to each user. Our mechanism admits an approximate ex-post NE. We also prove that it satisfies the joint differential privacy and is incentive-compatible. Efficiency of the approximate NE is evaluated, and the innovative scaling law results are disclosed. Finally, we provide simulation results to verify our analysis.

cs.GT↗

arXiv2016

Context-Aware Online Learning for Course Recommendation of MOOC Big Data

The Massive Open Online Course (MOOC) has expanded significantly in recent years. With the widespread of MOOC, the opportunity to study the fascinating courses for free has attracted numerous people of diverse educational backgrounds all over the world. In the big data era, a key research topic for MOOC is how to mine the needed courses in the massive course databases in cloud for each individual student accurately and rapidly as the number of courses is increasing fleetly. In this respect, the key challenge is how to realize personalized course recommendation as well as to reduce the computing and storage costs for the tremendous course data. In this paper, we propose a big data-supported, context-aware online learning-based course recommender system that could handle the dynamic and infinitely massive datasets, which recommends courses by using personalized context information and historical statistics. The context-awareness takes the personal preferences into consideration, making the recommendation suitable for people with different backgrounds. Besides, the algorithm achieves the sublinear regret performance, which means it can gradually recommend the mostly preferred and matched courses to students. In addition, our storage module is expanded to the distributed-connected storage nodes, where the devised algorithm can handle massive course storage problems from heterogeneous sources of course datasets. Comparing to existing algorithms, our proposed algorithms achieve the linear time complexity and space complexity. Experiment results verify the superiority of our algorithms when comparing with existing ones in the MOOC big data setting.

cs.LG↗

arXiv2016

Near Optimal Adaptive Shortest Path Routing with Stochastic Links States under Adversarial Attack

We consider the shortest path routing (SPR) of a network with stochastically time varying link metrics under potential adversarial attacks. Due to potential denial of service attacks, the distributions of link states could be stochastic (benign) or adversarial at different temporal and spatial locations. Without any \emph{a priori}, designing an adaptive SPR protocol to cope with all possible situations in practice optimally is a very challenging issue. In this paper, we present the first solution by formulating it as a multi-armed bandit (MAB) problem. By introducing a novel control parameter into the exploration phase for each link, a martingale inequality is applied in the our combinatorial adversarial MAB framework. As such, our proposed algorithms could automatically detect features of the environment within a unified framework and find the optimal SPR strategies with almost optimal learning performance in all possible cases over time. Moreover, we study important issues related to the practical implementation, such as decoupling route selection with multi-path route probing, cooperative learning among multiple sources, the "cold-start" issue and delayed feedback of our algorithm. Nonetheless, the proposed SPR algorithms can be implemented with low complexity and they are proved to scale very well with the network size. Comparing to existing approaches in a typical network scenario under jamming attacks, our algorithm has a 65.3\% improvement of network delay given a learning period and a 81.5\% improvement of learning duration under a specified network delay.

cs.NI↗

arXiv2016

Defect-free Perpendicular Diblock Copolymer Films: The Synergistic Effect of Surface Topography and Chemistry

We propose a direct self-assembly mechanism towards obtaining defect-free perpendicular lamellar phases of diblock copolymer (BCP) thin films. In our numerical study, a thin BCP film having a flat top surface is casted on a uni-directional corrugated solid substrate. The substrate is treated chemically and has a weak preference toward one of the two BCP components. Employing self-consistent field theory (SCFT), we find that there is an enhanced synergy between two substrate characteristics: its topography (geometrical roughness) combined with a weak surface preference. This synergy produces the desired perpendicular lamellar phase with perfect inplane ordering. Defect-free BCP lamellar phases are reproducible for several random initial states, and are obtained for a range of substrate roughness and chemical characteristics, even for a uni-directional multi-mode substrate roughness. Our theoretical study suggests possible experiments that will explore the interplay between uni-directional substrate corrugation and chemical surface treatment. It may lead to viable and economical ways of obtaining BCP films with defect-free lateral alignment.

cond-mat.soft↗

arXiv2016

On Diffusion-restricted Social Network: A Measurement Study of WeChat Moments

WeChat is a mobile messaging application that has 549 million active users as of Q1 2015, and "WeChat Moments" (WM) serves its social-networking function that allows users to post/share links of web pages. WM differs from the other social networks as it imposes many restrictions on the information diffusion process to mitigate the information overload. In this paper, we conduct a measurement study on information diffusion in the WM network by crawling and analyzing the spreading statistics of more than 160,000 pages that involve approximately 40 million users. Specifically, we identify the relationship of the number of posted pages and the number of views, the diffusion path length, the similarity and distribution of users' locations as well as their connections with the GDP of the users' province. For each individual WM page, we measure its temporal characteristics (e.g., the life time, the popularity within a time period); for each individual user, we evaluate how many of, or how likely, one's friends will view his posted pages. Our results will help the business to decide when and how to release the marketing pages over WM for better publicity.

cs.CY↗

arXiv2016

Distributed Private Online Learning for Social Big Data Computing over Data Center Networks

With the rapid growth of Internet technologies, cloud computing and social networks have become ubiquitous. An increasing number of people participate in social networks and massive online social data are obtained. In order to exploit knowledge from copious amounts of data obtained and predict social behavior of users, we urge to realize data mining in social networks. Almost all online websites use cloud services to effectively process the large scale of social data, which are gathered from distributed data centers. These data are so large-scale, high-dimension and widely distributed that we propose a distributed sparse online algorithm to handle them. Additionally, privacy-protection is an important point in social networks. We should not compromise the privacy of individuals in networks, while these social data are being learned for data mining. Thus we also consider the privacy problem in this article. Our simulations shows that the appropriate sparsity of data would enhance the performance of our algorithm and the privacy-preserving method does not significantly hurt the performance of the proposed algorithm.

cs.DC↗

arXiv2016

Secure Communication through Wireless-Powered Friendly Jamming: Jointly Online Optimization over Geography, Energy and Time

Exploring the interference-emitting friendly jammers to protect the sensitive communications in the presence of eavesdroppers has increasingly being investigated in literature. In parallel, scavenging energy from abient radio signals for energy-constrained devices, namely \emph{wireless energy harvesting} (WEH), has also drawn significant attention. Without relying on external energy supply, the wireless-powered friendly jammer by WEH from legitimate wireless devices is an effective approach to prolong their lifetime and gain the flexibility in deployments. This paper studies the online optimization of the placement and WEH of a set of friendly jammers in a geographic location with the energy-efficiency (EE) consideration. We adopt a simple "time switching" protocol where power transfer and jammer-assisted secure communications occur in different time blocks when WEH requests are launched. Our scheme has the following important advantages: 1) The proposed online jammers placement and interfering power allocation to attack eavesdroppers is the first distributed and scalable solutions within any specified geographic region, 2) We model the WEH for jammers as a \emph{JAM-NET lifetime maximization problem}, where online scheduling algorithms with heterogeneous energy demands of each jammer (from energy sources) are designed, 3) Under our model, the problem of placing a minimum number of jammers with distance-based power assignments is NP-hard, and near optimal PTAS approximation algorithms are provided, 4) When durations of the eavesdropping and legitimate communicating are available and the scenario is extended to the multi-channels setting, our results are strengthened to see further improved EE and reduced number of jammers. Simulations back up our theory.

cs.CR↗

arXiv2016

Differentially Private Online Learning for Cloud-Based Video Recommendation with Multimedia Big Data in Social Networks

With the rapid growth in multimedia services and the enormous offers of video contents in online social networks, users have difficulty in obtaining their interests. Therefore, various personalized recommendation systems have been proposed. However, they ignore that the accelerated proliferation of social media data has led to the big data era, which has greatly impeded the process of video recommendation. In addition, none of them has considered both the privacy of users' contexts (e,g., social status, ages and hobbies) and video service vendors' repositories, which are extremely sensitive and of significant commercial value. To handle the problems, we propose a cloud-assisted differentially private video recommendation system based on distributed online learning. In our framework, service vendors are modeled as distributed cooperative learners, recommending videos according to user's context, while simultaneously adapting the video-selection strategy based on user-click feedback to maximize total user clicks (reward). Considering the sparsity and heterogeneity of big social media data, we also propose a novel geometric differentially private model, which can greatly reduce the performance (recommendation accuracy) loss. Our simulation shows the proposed algorithms outperform other existing methods and keep a delicate balance between computing accuracy and privacy preserving level.

cs.LG↗

arXiv2016

Towards Optimal Adaptive Wireless Communications in Unknown Environments

Designing efficient channel access schemes for wireless communications without any prior knowledge about the nature of environments has been a very challenging issue, especially when the channel states distribution of all spectrum resources could be entirely or partially stochastic and/or adversarial at different time and locations. In this paper, we propose an adaptive channel access algorithm for wireless communications in unknown environments based on the theory of multi-armed bandits (MAB) problems. By automatically tuning two control parameters, i.e., learning rate and exploration probability, our algorithms are capable of finding the optimal channel access strategies and achieving the almost optimal learning performance over time under our defined four typical regimes for general unknown environments, e.g., the stochastic regime where channels follow some unknown i.i.d process, the adversarial regime where all channels are suffered by adversarial jamming attack, the mixed stochastic and adversarial regime where a subset of channels are attacked and the contaminated stochastic regime where occasionally adversarial events contaminate the stochastic channel process, etc. To reduce the implementation time and space complexity, we further develop an enhanced algorithm by exploiting the internal structure of the selection of channel access strategy. We conduct extensive simulations in all these regimes to validate our theoretical analysis. The quantitative performance studies indicate the superior throughput gain and the flexibility of our algorithm in practice, which is resilient to both oblivious and adaptive jamming attacks with different intelligence and any attacking strength that ranges from no-attack to the full-attack of all spectrum resources.

cs.NI↗

arXiv2015

Lamellar Diblock Copolymers on Rough Substrates: Self-consistent Field Theory Studies

We present numerical calculations of lamellar phases of di-block copolymers (BCP) confined between two surfaces, where the top surface is flat and the bottom one is corrugated. The corrugated substrate is assumed to have a single $q$-mode of lateral undulations with a wavenumber q_s and amplitude R. We focus on the effects of substrate roughness, parameterized by the dimensionless quantity, q_sR, on the relative stability between parallel and perpendicular orientations of the lamellar phase. The competition between film confinement, energy cost of elastic deformation and gain in surface energy induces a parallel-to-perpendicular transition of the BCP lamellae. Employing self-consistent field theory (SCFT), we study the critical substrate roughness value corresponding to this transition. The critical value increases as function of the surface preference towards one of the two BCP components, and as function of film thickness. But, it decreases with increasing values of the Flory-Huggins parameter. Our findings are equivalent to stating that the critical value decreases as the BCP molecular weight or the natural BCP periodicity increases. We further show that the rough substrate can overcome the formation of parallel lamellae in cases where the top surface has a preference towards one of the two BCP components. Our results are in good agreement with previous experiments, and highlight the physical conditions behind the perpendicular orientation of lamellar phases, as is desired in nanolithography and other industrial applications.

cond-mat.soft↗

arXiv2015

Almost Optimal Energy-Efficient Cognitive Communications in Unknown Environments

Cognitive (Radio) (CR) Communications (CC) are mainly deployed within the environments of primary (user) communications, where the channel states and accessibility are usually stochastically distributed (benign or IID). However, many practical CC are also exposed to disturbing events (contaminated) and vulnerable jamming attacks (adversarial or non-IID). Thus, the channel state distribution of spectrum could be stochastic, contaminated or adversarial at different temporal and spatial locations. Without any a priori, facilitating optimal CC is a very challenging issue. In this paper, we propose an online learning algorithm that performs the joint channel sensing, probing and adaptive channel access for multi-channel CC in general unknown environments. We take energy-efficient CC (EECC) into our special attention, which is highly desirable for green wireless communications and demanding to combat with potential jamming attack who could greatly mar the energy and spectrum efficiency of CC. The EECC is formulated as a constrained regret minimization problem with power budget constraints. By tuning a novel exploration parameter, our algorithms could adaptively find the optimal channel access strategies and achieve the almost optimal learning performance of EECC in different scenarios provided with the vanishing long-term power budget violations. We also consider the important scenario that cooperative learning and information sharing among multiple CR users to see further performance improvements. The proposed algorithms are resilient to both oblivious and adaptive jamming attacks with different intelligence and attacking strength. Extensive numerical results are conducted to validate our theory.

cs.NI↗

arXiv2015

Differentially Private Distributed Online Learning

Online learning has been in the spotlight from the machine learning society for a long time. To handle massive data in Big Data era, one single learner could never efficiently finish this heavy task. Hence, in this paper, we propose a novel distributed online learning algorithm to solve the problem. Comparing to typical centralized online learner, the distributed learners optimize their own learning parameters based on local data sources and timely communicate with neighbors. However, communication may lead to a privacy breach. Thus, we use differential privacy to preserve the privacy of learners, and study the influence of guaranteeing differential privacy on the utility of the distributed online learning algorithm. Furthermore, by using the results from Kakade and Tewari (2009), we use the regret bounds of online learning to achieve fast convergence rates for offline learning algorithms in distributed scenarios, which provides tighter utility performance than the existing state-of-the-art results. In simulation, we demonstrate that the differentially private offline learning algorithm has high variance, but we can use mini-batch to improve the performance. Finally, the simulations show that the analytical results of our proposed theorems are right and our private distributed online learning algorithm is a general framework.

cs.LG↗

arXiv2013

Magnetic Exchange Coupling and Anisotropy of 3d Transition-Metal Nanowire on the Surface of Graphyne Sheet

Using density functional theory plus Hubbard-U (DFT+U) approach, we find that quasi one-dementation(1D) 3d transition metal(TM) zigzag nanowire can be constructed by TM adsorbed on the surface of graphyne sheet. The results show that the TM exchange coupling of the zigzag nanowire mediated by sp hybridized carbon atoms gives rise to long range ferromagnetic order except for Cr with anti-ferromagnetic order. The magnetic exchange interaction of TM chains follows like-Zener's p_z-d exchange mechanism: the coexistence of out-of plane p_z-d and in-plane p_x-y-d exchange. Finally, by including spin-orbit interactions within spin-DFT, we calculate the magnetic anisotropy energy of the TM chain on graphyne. We find that the Fe and Co chains show considerable magnetic anisotropy energy (MAE) and orbital magnetic moment. The easy axis of V, Cr, Mn and Fe chains is perpendicular to the surface, whereas the easy axis of Co lies in the surface. Moreover, only V chain shows relatively larger in-plane anisotropy. Our results open a new route to realize the applications of graphyne in spintronics.

cond-mat.mtrl-sci↗

arXiv2013

Magnetic Properties of Single Transition-Metal Atom Absorbed Graphdiyne and Graphyne Sheet

The electronic and magnetic properties of single 3d transition-metal(TM) atom (V, Cr, Mn, Fe, Co, and Ni) adsorbed graphdiyne (GDY) and graphyne (GY) are systematically studied using first-principles calculations within the density functional framework. We find that the adsorption of TM atom not only efficiently modulates the electronic structures of GDY/GY system, but also introduces excellent magnetic properties, such as half-metal and spin-select half-semiconductor. Such modulation originates from the charge transfer between TM adatom and the GDY/GY sheet as well as the electron redistribution of the TM intra-atomic s, p, and d orbitals. Our results indicate that the TM adsorbed GDY/GY are excellent candidates for spintronics.

cond-mat.mtrl-sci↗