arXiv subjects

Tom Z. J. Fu

Publications and source records attributed to Tom Z. J. Fu.

11 recordsLinked to original sources

arXiv2020

DiffPerf: Towards Performance Differentiation and Optimization with SDN Implementation

Continuing the current trend, Internet traffic is expected to grow significantly over the coming years, with video traffic consuming the biggest share. On the one hand, this growth poses challenges to access providers (APs), who have to upgrade their infrastructure to meet the growing traffic demands as well as find new ways to monetize their network resources. On the other hand, despite numerous optimizations of the underlying transport protocol, a user's utilization of network bandwidth and is thus the user's perceived quality still being largely affected by network latency and buffer size. To address both concerns, we propose DiffPerf, a class-based differentiation framework, that, at a macroscopic level dynamically allocates bandwidth to service classes pre-defined by the APs, and at a microscopic level statistically differentiates and isolates user flows to help them achieve better performance. We implement DiffPerf on OpenDaylight SDN controller and programmable Barefoot Tofino switch and evaluate it from an application perspective for MPEG-DASH video streaming. Our evaluations demonstrate the practicality and flexibility that DiffPerf provides APs with capabilities through which a spectrum of qualities are provisioned at multiple classes. Meanwhile, it assists in achieving better fairness and improving overall user's perceived quality within the same class.

cs.NI↗

arXiv2019

Causal Mechanism Transfer Network for Time Series Domain Adaptation in Mechanical Systems

Data-driven models are becoming essential parts in modern mechanical systems, commonly used to capture the behavior of various equipment and varying environmental characteristics. Despite the advantages of these data-driven models on excellent adaptivity to high dynamics and aging equipment, they are usually hungry to massive labels over historical data, mostly contributed by human engineers at an extremely high cost. The label demand is now the major limiting factor to modeling accuracy, hindering the fulfillment of visions for applications. Fortunately, domain adaptation enhances the model generalization by utilizing the labelled source data as well as the unlabelled target data and then we can reuse the model on different domains. However, the mainstream domain adaptation methods cannot achieve ideal performance on time series data, because most of them focus on static samples and even the existing time series domain adaptation methods ignore the properties of time series data, such as temporal causal mechanism. In this paper, we assume that causal mechanism is invariant and present our Causal Mechanism Transfer Network(CMTN) for time series domain adaptation. By capturing and transferring the dynamic and temporal causal mechanism of multivariate time series data and alleviating the time lags and different value ranges among different machines, CMTN allows the data-driven models to exploit existing data and labels from similar systems, such that the resulting model on a new system is highly reliable even with very limited data. We report our empirical results and lessons learned from two real-world case studies, on chiller plant energy optimization and boiler fault detection, which outperforms the existing state-of-the-art method.

cs.LG↗

arXiv2019

On SDN-Enabled Online and Dynamic Bandwidth Allocation for Stream Analytics

Data communication in cloud-based distributed stream data analytics often involves a collection of parallel and pipelined TCP flows. As the standard TCP congestion control mechanism is designed for achieving "fairness" among competing flows and is agnostic to the application layer contexts, the bandwidth allocation among a set of TCP flows traversing bottleneck links often leads to sub-optimal application-layer performance measures, e.g., stream processing throughput or average tuple complete latency. Motivated by this and enabled by the rapid development of the Software-Defined Networking (SDN) techniques, in this paper, we re-investigate the design space of the bandwidth allocation problem and propose a cross-layer framework which utilizes the additional information obtained from the application layer and provides on-the-fly and dynamic bandwidth adjustment algorithms for helping the stream analytics applications achieving better performance during the runtime. We implement a prototype cross-layer bandwidth allocation framework based on a popular open-source distributed stream processing platform, Apache Storm, together with the OpenDaylight controller, and carry out extensive experiments with real-world analytical workloads on top of a local cluster consisting of 10 workstations interconnected by a SDN-enabled switch. The experiment results clearly validate the effectiveness and efficiency of our proposed framework and algorithms.

cs.NI↗

arXiv2017

Elasticutor: Rapid Elasticity for Realtime Stateful Stream Processing

Elasticity is highly desirable for stream processing systems to guarantee low latency against workload dynamics, such as surges in data arrival rate and fluctuations in data distribution. Existing systems achieve elasticity following a resource-centric approach that uses dynamic key partitioning across the parallel instances, i.e. executors, to balance the workload and scale operators. However, such operator-level key repartitioning needs global synchronization and prohibits rapid elasticity. To address this problem, we propose an executor-centric approach, whose core idea is to avoid operator-level key repartitioning while implementing each executor as the building block of elasticity. Following this new approach, we design the Elasticutor framework with two level of optimizations: i) a novel implementation of executors, i.e., elastic executors, that perform elastic multi-core execution via efficient intra-executor load balancing and executor scaling and ii) a global model-based scheduler that dynamically allocates CPU cores to executors based on the instantaneous workloads. We implemented a prototype of Elasticutor and conducted extensive experiments. Our results show that Elasticutor doubles the throughput and achieves an average processing latency up to 2 orders of magnitude lower than previous methods, for a dynamic workload of real-world applications.

cs.DB↗

arXiv2017

Distributed Publish/Subscribe Query Processing on the Spatio-Textual Data Stream

Huge amount of data with both space and text information, e.g., geo-tagged tweets, is flooding on the Internet. Such spatio-textual data stream contains valuable information for millions of users with various interests on different keywords and locations. Publish/subscribe systems enable efficient and effective information distribution by allowing users to register continuous queries with both spatial and textual constraints. However, the explosive growth of data scale and user base has posed challenges to the existing centralized publish/subscribe systems for spatio-textual data streams. In this paper, we propose our distributed publish/subscribe system, called PS2Stream, which digests a massive spatio-textual data stream and directs the stream to target users with registered interests. Compared with existing systems, PS2Stream achieves a better workload distribution in terms of both minimizing the total amount of workload and balancing the load of workers. To achieve this, we propose a new workload distribution algorithm considering both space and text properties of the data. Additionally, PS2Stream supports dynamic load adjustments to adapt to the change of the workload, which makes PS2Stream adaptive. Extensive empirical evaluation, on commercial cloud computing platform with real data, validates the superiority of our system design and advantages of our techniques on system performance improvement.

cs.DB↗

arXiv2016

Parallel Stream Processing Against Workload Skewness and Variance

Key-based workload partitioning is a common strategy used in parallel stream processing engines, enabling effective key-value tuple distribution over worker threads in a logical operator. While randomized hashing on the keys is capable of balancing the workload for key-based partitioning when the keys generally follow a static distribution, it is likely to generate poor balancing performance when workload variance occurs on the incoming data stream. This paper presents a new key-based workload partitioning framework, with practical algorithms to support dynamic workload assignment for stateful operators. The framework combines hash-based and explicit key-based routing strategies for workload distribution, which specifies the destination worker threads for a handful of keys and assigns the other keys with the hashing function. When short-term distribution fluctuations occur to the incoming data stream, the system adaptively updates the routing table containing the chosen keys, in order to rebalance the workload with minimal migration overhead within the stateful operator. We formulate the rebalance operation as an optimization problem, with multiple objectives on minimizing state migration costs, controlling the size of the routing table and breaking workload imbalance among worker threads. Despite of the NP-hardness nature behind the optimization formulation, we carefully investigate and justify the heuristics behind key (re)routing and state migration, to facilitate fast response to workload variance with ignorable cost to the normal processing in the distributed system. Empirical studies on synthetic data and real-world stream applications validate the usefulness of our proposals and prove the huge advantage of our approaches over state-of-the-art solutions in the literature.

cs.DC↗

arXiv2016

Component-Based Distributed Framework for Coherent and Real-Time Video Dehazing

Traditional dehazing techniques, as a well studied topic in image processing, are now widely used to eliminate the haze effects from individual images. However, even the state-of-the-art dehazing algorithms may not provide sufficient support to video analytics, as a crucial pre-processing step for video-based decision making systems (e.g., robot navigation), due to the limitations of these algorithms on poor result coherence and low processing efficiency. This paper presents a new framework, particularly designed for video dehazing, to output coherent results in real time, with two novel techniques. Firstly, we decompose the dehazing algorithms into three generic components, namely transmission map estimator, atmospheric light estimator and haze-free image generator. They can be simultaneously processed by multiple threads in the distributed system, such that the processing efficiency is optimized by automatic CPU resource allocation based on the workloads. Secondly, a cross-frame normalization scheme is proposed to enhance the coherence among consecutive frames, by sharing the parameters of atmospheric light from consecutive frames in the distributed computation platform. The combination of these techniques enables our framework to generate highly consistent and accurate dehazing results in real-time, by using only 3 PCs connected by Ethernet.

cs.DC↗

arXiv2015

Optimal Operator State Migration for Elastic Data Stream Processing

A cloud-based data stream management system (DSMS) handles fast data by utilizing the massively parallel processing capabilities of the underlying platform. An important property of such a DSMS is elasticity, meaning that nodes can be dynamically added to or removed from an application to match the latter's workload, which may fluctuate in an unpredictable manner. For an application involving stateful operations such as aggregates, the addition / removal of nodes necessitates the migration of operator states. Although the importance of migration has been recognized in existing systems, two key problems remain largely neglected, namely how to migrate and what to migrate, i.e., the migration mechanism that reduces synchronization overhead and result delay during migration, and the selection of the optimal task assignment that minimizes migration costs. Consequently, migration in current systems typically incurs a high spike in result delay caused by expensive synchronization barriers and suboptimal task assignments. Motivated by this, we present the first comprehensive study on efficient operator states migration, and propose designs and algorithms that enable live, progressive, and optimized migrations. Extensive experiments using real data justify our performance claims.

cs.DC↗

arXiv2015

DRS: Dynamic Resource Scheduling for Real-Time Analytics over Fast Streams

In a data stream management system (DSMS), users register continuous queries, and receive result updates as data arrive and expire. We focus on applications with real-time constraints, in which the user must receive each result update within a given period after the update occurs. To handle fast data, the DSMS is commonly placed on top of a cloud infrastructure. Because stream properties such as arrival rates can fluctuate unpredictably, cloud resources must be dynamically provisioned and scheduled accordingly to ensure real-time response. It is quite essential, for the existing systems or future developments, to possess the ability of scheduling resources dynamically according to the current workload, in order to avoid wasting resources, or failing in delivering correct results on time. Motivated by this, we propose DRS, a novel dynamic resource scheduler for cloud-based DSMSs. DRS overcomes three fundamental challenges: (a) how to model the relationship between the provisioned resources and query response time (b) where to best place resources; and (c) how to measure system load with minimal overhead. In particular, DRS includes an accurate performance model based on the theory of \emph{Jackson open queueing networks} and is capable of handling \emph{arbitrary} operator topologies, possibly with loops, splits and joins. Extensive experiments with real data confirm that DRS achieves real-time response with close to optimal resource consumption.

cs.DC↗

arXiv2014

MYE: Missing Year Estimation in Academic Social Networks

In bibliometrics studies, a common challenge is how to deal with incorrect or incomplete data. However, given a large volume of data, there often exists certain relationships between the data items that can allow us to recover missing data items and correct erroneous data. In this paper, we study a particular problem of this sort - estimating the missing year information associated with publications (and hence authors' years of active publication). We first propose a simple algorithm that only makes use of the "direct" information, such as paper citation/reference relationships or paper-author relationships. The result of this simple algorithm is used as a benchmark for comparison. Our goal is to develop algorithms that increase both the coverage (the percentage of missing year papers recovered) and accuracy (mean absolute error of the estimated year to the real year). We propose some advanced algorithms that extend inference by information propagation. For each algorithm, we propose three versions according to the given academic social network type: a) Homogeneous (only contains paper citation links), b) Bipartite (only contains paper-author relations), and, c) Heterogeneous (both paper citation and paper-author relations). We carry out experiments on the three public data sets (MSR Libra, DBLP and APS), and evaluated by applying the K-fold cross validation method. We show that the advanced algorithms can improve both coverage and accuracy.

cs.DL↗

arXiv2014

The Academic Social Network

Through academic publications, the authors of these publications form a social network. Instead of sharing casual thoughts and photos (as in Facebook), authors pick co-authors and reference papers written by other authors. Thanks to various efforts (such as Microsoft Libra and DBLP), the data necessary for analyzing the academic social network is becoming more available on the Internet. What type of information and queries would be useful for users to find out, beyond the search queries already available from services such as Google Scholar? In this paper, we explore this question by defining a variety of ranking metrics on different entities -authors, publication venues and institutions. We go beyond traditional metrics such as paper counts, citations and h-index. Specifically, we define metrics such as influence, connections and exposure for authors. An author gains influence by receiving more citations, but also citations from influential authors. An author increases his/her connections by co-authoring with other authors, and specially from other authors with high connections. An author receives exposure by publishing in selective venues where publications received high citations in the past, and the selectivity of these venues also depends on the influence of the authors who publish there. We discuss the computation aspects of these metrics, and similarity between different metrics. With additional information of author-institution relationships, we are able to study institution rankings based on the corresponding authors' rankings for each type of metric as well as different domains. We are prepared to demonstrate these ideas with a web site (http://pubstat.org) built from millions of publications and authors.

cs.SI↗