arXiv subjects

Dan Li

Publications and source records attributed to Dan Li.

At least 181 records · Page 10Linked to original sources

arXiv2021

Analyzing Open-Source Serverless Platforms: Characteristics and Performance

Serverless computing is increasingly popular because of its lower cost and easier deployment. Several cloud service providers (CSPs) offer serverless computing on their public clouds, but it may bring the vendor lock-in risk. To avoid this limitation, many open-source serverless platforms come out to allow developers to freely deploy and manage functions on self-hosted clouds. However, building effective functions requires much expertise and thorough comprehension of platform frameworks and features that affect performance. It is a challenge for a service developer to differentiate and select the appropriate serverless platform for different demands and scenarios. Thus, we elaborate the frameworks and event processing models of four popular open-source serverless platforms and identify their salient idiosyncrasies. We analyze the root causes of performance differences between different service exporting and auto-scaling modes on those platforms. Further, we provide several insights for future work, such as auto-scaling and metric collection.

cs.DC↗

arXiv2021

High Performance and Scalable NAT System on Commodity Platforms

Quick network address translation (NAT) is proposed to improve the network performance of the NAT system on the commodity server by three ways. First, the quick NAT search algorithm is designed to use the Hash search instead of the sequential search to reduce latency when looking up the NAT rule table. Second, to leverage the power of the multi-core central processing unit (CPU) and the multi-queue network interface card, Quick NAT enables multiple CPU cores to process in parallel. The localized connection tracking table and the compare-and-swap based lock-free NAT Hash tables are designed to eliminate the lock overhead. Third, Quick NAT uses the polling and zero-copy delivery to reduce the cost of interrupt and packet copies. The evaluation results show that Quick NAT obtains high scalability and line-rate throughput on the commodity server.

cs.NI↗

arXiv2021

Quick NAT: High performance NAT system on commodity platforms

NAT gateway is an important network system in today's IPv4 network when translating a private IPv4 address to a public address. However, traditional NAT system based on Linux Netfilter cannot achieve high network throughput to meet modern requirements such as data centers. To address this challenge, we improve the network performance of NAT system by three ways. First, we leverage DPDK to enable polling and zero-copy delivery, so as to reduce the cost of interrupt and packet copies. Second, we enable multiple CPU cores to process in parallel and use lock-free hash table to minimize the contention between CPU cores. Third, we use hash search instead of sequential search when looking up the NAT rule table. Evaluation shows that our Quick NAT system significantly improves the performance of NAT on commodity platforms.

cs.NI↗

arXiv2021

Data-driven Predictive Control for a Class of Uncertain Control-Affine Systems

This paper studies a data-driven predictive control for a class of control-affine systems which is subject to uncertainty. With the accessibility to finite sample measurements of the uncertain variables, we aim to find controls which are feasible and provide superior performance guarantees with high probability. This results into the formulation of a stochastic optimization problem (P), which is intractable due to the unknown distribution of the uncertainty variables. By developing a distributionally robust optimization framework, we present an equivalent and yet tractable reformulation of (P). Further, we propose an efficient algorithm that provides online suboptimal data-driven solutions and guarantees performance with high probability. To illustrate the effectiveness of the proposed approach, we consider a highway speed-limit control problem. We then develop a set of data-driven speed controls that allow us to prevent traffic congestion with high probability. Finally, we employ the resulting control method on a traffic simulator to illustrate the effectiveness of this approach numerically.

math.OC↗

arXiv2021

Exciton emissions in bilayer WSe2 tuned by the ferroelectric polymer

In this work, a hybrid integration of few-layer transition metal dichalcogenides (TMDCs) and ferroelectric polymer is designed to achieve passive control of optical properties in-situ. The electrical polarization in ferroelectric P(VDF-TrFE) polymer can regulate the photoluminescence (PL) in few-layer TMDCs. The total PL intensity is substantially suppressed or enhanced under opposite polarization in bilayer WSe2. This is because electrons transfer between valley K and Λ in the conduction band induced by the built-in electric field in P(VDF-TrFE) polymer. This charge transfer further changes the competing dynamics between direct and indirect exciton recombination path and overall optical radiation efficiency. We also illustrate that the engineered PL originates from external electric field dependent transferred electron effect. The theoretical result matches the experimental data well. This work demonstrates a device platform in which passive regulation is achieved using 2D TMDCs modulated by polarized ferroelectric materials.

physics.optics↗

arXiv2021

PC-GAIN: Pseudo-label Conditional Generative Adversarial Imputation Networks for Incomplete Data

Datasets with missing values are very common in real world applications. GAIN, a recently proposed deep generative model for missing data imputation, has been proved to outperform many state-of-the-art methods. But GAIN only uses a reconstruction loss in the generator to minimize the imputation error of the non-missing part, ignoring the potential category information which can reflect the relationship between samples. In this paper, we propose a novel unsupervised missing data imputation method named PC-GAIN, which utilizes potential category information to further enhance the imputation power. Specifically, we first propose a pre-training procedure to learn potential category information contained in a subset of low-missing-rate data. Then an auxiliary classifier is determined using the synthetic pseudo-labels. Further, this classifier is incorporated into the generative adversarial framework to help the generator to yield higher quality imputation results. The proposed method can improve the imputation quality of GAIN significantly. Experimental results on various benchmark datasets show that our method is also superior to other baseline approaches. Our code is available at \url{https://github.com/WYu-Feng/pc-gain}.

cs.LG↗

arXiv2021

Dynamical evolution of the inner asteroid belt

A determination of the dynamical evolution of the asteroid belt is difficult because the asteroid belt has evolved since the time of asteroid formation through mechanisms that include: (1) catastrophic collisions, (2) rotational disruption, (3) chaotic orbital evolution and (4) orbital evolution driven by Yarkovsky radiation forces. The timescales of these loss mechanisms are uncertain and there is a need for more observational constraints. In the inner main belt, the mean size of the non-family asteroids increases with increasing inclination. Here, we use that observation to show that all inner main belt asteroids originate from either the known families or from ghost families, that is, old families with dispersed orbital elements. We estimate that the average age of the asteroids in the ghost families is a factor of 1/3 less than the Yarkovsky orbital evolution timescale. However, this orbital evolution timescale is a long-term average that must allow for the collisional evolution of the asteroids and for stochastic changes in their spin directions. By applying these constraints on the orbital evolution timescales to the evolution of the size-frequency distribution of the Vesta asteroid family, we estimate that the age of this family is greater than 1.3 $Gyr$ and could be comparable with the age of the solar system. By estimating the number of ghost families, we calculate that the number of asteroids that are the root sources of the meteorites and the near-Earth asteroids that originate from the inner main belt is about 20.

astro-ph.EP↗

arXiv2021

An Online Approach to Cyberattack Detection and Localization in Smart Grid

Complex interconnections between information technology and digital control systems have significantly increased cybersecurity vulnerabilities in smart grids. Cyberattacks involving data integrity can be very disruptive because of their potential to compromise physical control by manipulating measurement data. This is especially true in large and complex electric networks that often rely on traditional intrusion detection systems focused on monitoring network traffic. In this paper, we develop an online detection algorithm to detect and localize covert attacks on smart grids. Using a network system model, we develop a theoretical framework by characterizing a covert attack on a generator bus in the network as sparse features in the state-estimation residuals. We leverage such sparsity via a regularized linear regression method to detect and localize covert attacks based on the regression coefficients. We conduct a comprehensive numerical study on both linear and nonlinear system models to validate our proposed method. The results show that our method outperforms conventional methods in both detection delay and localization accuracy.

stat.AP↗

arXiv2021

Electrocardiogram Classification and Visual Diagnosis of Atrial Fibrillation with DenseECG

Atrial Fibrillation (AF) is a common cardiac arrhythmia affecting a large number of people around the world. If left undetected, it will develop into chronic disability or even early mortality. However, patients who have this problem can barely feel its presence, especially in its early stage. A non-invasive, automatic, and effective detection method is therefore needed to help early detection so that medical intervention can be implemented in time to prevent its progression. Electrocardiogram (ECG), which records the electrical activities of the heart, has been widely used for detecting the presence of AF. However, due to the subtle patterns of AF, the performance of detection models have largely depended on complicated data pre-processing and expertly engineered features. In our work, we developed DenseECG, an end-to-end model based on 5 layers 1D densely connected convolutional neural network. We trained our model using the publicly available dataset from 2017 PhysioNet Computing in Cardiology(CinC) Challenge containing 8528 single-lead ECG recordings of short-term heart rhythms (9-61s). Our trained model was able to outperform the other state-of-the-art AF detection models on this dataset without complicated data pre-processing and expert-supervised feature engineering.

eess.SP↗

arXiv2020

The common origin of family and non-family asteroids

All asteroids are currently classified as either family, originating from the disruption of known bodies, or non-family. An outstanding question is the origin of these non-family asteroids. Were they formed individually, or as members of known families but with chaotically evolving orbits, or are they members of old ghost families, that is, asteroids with a common parent body but with orbits that no longer cluster in orbital element space? Here, we show that the sizes of the non-family asteroids in the inner belt are correlated with their orbital eccentricities and anticorrelated with their inclinations, suggesting that both non-family and family asteroids originate from a small number of large primordial planetesimals. We estimate that ~85% of the asteroids in the inner main belt originate from the Flora, Vesta, Nysa, Polana and Eulalia families, with the remaining ~15% originating from either the same families or, more likely, a few ghost families. These new results imply that we must seek explanations for the differing characteristics of the various meteorite groups in the evolutionary histories of a few, large, precursor bodies. Our findings also support the model that asteroids formed big through the gravitational collapse of material in a protoplanetary disk.

astro-ph.EP↗

arXiv2020

Improve Adversarial Robustness via Weight Penalization on Classification Layer

It is well-known that deep neural networks are vulnerable to adversarial attacks. Recent studies show that well-designed classification parts can lead to better robustness. However, there is still much space for improvement along this line. In this paper, we first prove that, from a geometric point of view, the robustness of a neural network is equivalent to some angular margin condition of the classifier weights. We then explain why ReLU type function is not a good choice for activation under this framework. These findings reveal the limitations of the existing approaches and lead us to develop a novel light-weight-penalized defensive method, which is simple and has a good scalability. Empirical results on multiple benchmark datasets demonstrate that our method can effectively improve the robustness of the network without requiring too much additional computation, while maintaining a high classification precision for clean data.

cs.LG↗

arXiv2020

Deep Learning based Covert Attack Identification for Industrial Control Systems

Cybersecurity of Industrial Control Systems (ICS) is drawing significant concerns as data communication increasingly leverages wireless networks. A lot of data-driven methods were developed for detecting cyberattacks, but few are focused on distinguishing them from equipment faults. In this paper, we develop a data-driven framework that can be used to detect, diagnose, and localize a type of cyberattack called covert attacks on smart grids. The framework has a hybrid design that combines an autoencoder, a recurrent neural network (RNN) with a Long-Short-Term-Memory (LSTM) layer, and a Deep Neural Network (DNN). This data-driven framework considers the temporal behavior of a generic physical system that extracts features from the time series of the sensor measurements that can be used for detecting covert attacks, distinguishing them from equipment faults, as well as localize the attack/fault. We evaluate the performance of the proposed method through a realistic simulation study on the IEEE 14-bus model as a typical example of ICS. We compare the performance of the proposed method with the traditional model-based method to show its applicability and efficacy.

cs.LG↗

arXiv2020

Online Learning of Parameterized Uncertain Dynamical Environments with Finite-sample Guarantees

We present a novel online learning algorithm for a class of unknown and uncertain dynamical environments that are fully observable. First, we obtain a novel probabilistic characterization of systems whose mean behavior is known but which are subject to additive, unknown subGaussian disturbances. This characterization relies on recent concentration of measure results and is given in terms of ambiguity sets. Second, we extend the results to environments whose mean behavior is also unknown but described by a parameterized class of possible mean behaviors. Our algorithm adapts the ambiguity set dynamically by learning the parametric dependence online, and retaining similar probabilistic guarantees with respect to the additive, unknown disturbance. We illustrate the results on a differential-drive robot subject to environmental uncertainty.

eess.SY↗

arXiv2020

Grover search with smaller oracles

Grover search is one of the most important quantum algorithms. In this paper, we consider a kind of search that the conditions of satisfaction $T$ can be rewritten as $T=T_1\bigcap T_2$. Then we present a new Grover search with smaller oracles. The time complexity of this algorithm $O(\fracπ{4}\sqrt{\frac{N}{bλ}}+\fracπ{4}\sqrt{\frac{b}τ})$, which is smaller than the time complexity of original Grover search, i.e. $O(\fracπ{4}\sqrt{\frac{N}{M}})$.

physics.gen-ph↗

arXiv2020

Data assimilation and online optimization with performance guarantees

This paper considers a class of real-time stochastic optimization problems dependent on an unknown probability distribution. In the considered scenario, data is streaming frequently while trying to reach a decision. Thus, we aim to devise a procedure that incorporates samples (data) of the distribution sequentially and adjusts decisions accordingly. We approach this problem in a distributionally robust optimization framework and propose a novel Online Data Assimilation Algorithm (ONDA Algorithm) for this purpose. This algorithm guarantees out-of-sample performance of decisions with high probability, and gradually improves the quality of the decisions by incorporating the streaming data. We show that the ONDA Algorithm converges under a sufficiently slow data streaming rate, and provide a criteria for its termination after certain number of data have been collected. Simulations illustrate the results.

math.OC↗

arXiv2020

A simplified primal-dual weak Galerkin finite element method for Fokker-Planck type equations

A simplified primal-dual weak Galerkin (S-PDWG) finite element method is designed for the Fokker-Planck type equation with non-smooth diffusion tensor and drift vector. The discrete system resulting from S-PDWG method has significantly fewer degrees of freedom compared with the one resulting from the PDWG method proposed by Wang-Wang \cite{WW-fp-2018}. Furthermore, the condition number of the S-PDWG method is smaller than the PDWG method \cite{WW-fp-2018} due to the introduction of a new stabilizer, which provides a potential for designing fast algorithms. Optimal order error estimates for the S-PDWG approximation are established in the $L^2$ norm. A series of numerical results are demonstrated to validate the effectiveness of the S-PDWG method.

math.NA↗

arXiv2020

Three-state quantum walk on the Cayley Graph of the Dihedral Group

The finite dihedral group generated by one rotation and one reflection is the simplest case of the non-abelian group. Cayley graphs are diagrammatic counterparts of groups. In this paper, much attention is given to the Cayley graph of the dihedral group. Considering the characteristics of the elements in the dihedral group, we propose a model of three-state discrete-time quantum walk (DTQW) on the Caylay graph of the dihedral group with Grover coin. We derive analytic expressions for the the position probability distribution and the long-time limit of the return probability starting from the origin. It is shown that the localization effect is governed by the size of the underlying dihedral group, coin operator and initial state. We also numerically investigate the properties of the proposed model via the probability distribution and the time-averaged probability at the designated position. The abundant phenomena of three-state Grover DTQW on the Caylay graph of the dihedral group can help the community to better understand and to develop new quantum algorithms.

quant-ph↗

arXiv2020

Query Resolution for Conversational Search with Limited Supervision

In this work we focus on multi-turn passage retrieval as a crucial component of conversational search. One of the key challenges in multi-turn passage retrieval comes from the fact that the current turn query is often underspecified due to zero anaphora, topic change, or topic return. Context from the conversational history can be used to arrive at a better expression of the current turn query, defined as the task of query resolution. In this paper, we model the query resolution task as a binary term classification problem: for each term appearing in the previous turns of the conversation decide whether to add it to the current turn query or not. We propose QuReTeC (Query Resolution by Term Classification), a neural query resolution model based on bidirectional transformers. We propose a distant supervision method to automatically generate training data by using query-passage relevance labels. Such labels are often readily available in a collection either as human annotations or inferred from user interactions. We show that QuReTeC outperforms state-of-the-art models, and furthermore, that our distant supervision method can be used to substantially reduce the amount of human-curated data required to train QuReTeC. We incorporate QuReTeC in a multi-turn, multi-stage passage retrieval architecture and demonstrate its effectiveness on the TREC CAsT dataset.

cs.IR↗