Cloud computing has been bringing fundamental changes to computing models in the past few years. Infrastructure as a Service (IaaS), a kind of basic cloud services, is provisioned to customers in the form of virtual machines (VMs). The increasing demands for IaaS cloud services require the performability analysis of cloud infrastructure. Analytic modeling is one of the effective evaluation approaches. This paper aims to develop a monolithic model, by using continuous time Markov chain (CTMC), for a IaaS CDC, which (1) consists of active and standby physical machines (PMs), (2) allows PM migration among active and standby PM pools, (3) all jobs are homogeneous, and (4) a running job could continue its running by using idle active PMs when the PM working for this job fails. Although a monolithic CTMC model for IaaS Cloud performability analysis may face largeness and stiffness problems, it could be used to verify the scalable approximate model. We present the details of state transition rules of the proposed model and the formula for computing metrics, including the immediate service probability, the mean response time and so on. Numerical analysis and simulations are carried out to verify the accuracy of the proposed model.
在过去的几年里，云计算已经给计算模型带来了根本性的变化。IaaS (Infrastructure as a Service)是一种基础云服务，以虚拟机(vm)的形式提供给客户。对IaaS云服务日益增长的需求要求对云基础设施进行性能分析。分析建模是一种有效的评价方法。本文旨在通过使用连续时间马尔可夫链(CTMC)为IaaS CDC开发一个整体模型，该模型(1)由主备物理机(PM)组成，(2)允许PM在主备PM池之间迁移，(3)所有作业都是同构的，(4)当为该作业工作的PM失败时，运行中的作业可以通过使用空闲的活动PM继续运行。尽管用于IaaS云性能分析的单片CTMC模型可能面临较大和刚度问题，但它可以用于验证可扩展的近似模型。给出了该模型的状态转移规则的详细信息，并给出了计算指标的公式，包括即时服务概率、平均响应时间等。通过数值分析和仿真验证了所提模型的准确性。
{"title":"Performability Analysis for IaaS Cloud Data Center","authors":"T. Wang, Xiaolin Chang, Bo Liu","doi":"10.1109/PDCAT.2016.033","DOIUrl":"https://doi.org/10.1109/PDCAT.2016.033","url":null,"abstract":"Cloud computing has been bringing fundamental changes to computing models in the past few years. Infrastructure as a Service (IaaS), a kind of basic cloud services, is provisioned to customers in the form of virtual machines (VMs). The increasing demands for IaaS cloud services require the performability analysis of cloud infrastructure. Analytic modeling is one of the effective evaluation approaches. This paper aims to develop a monolithic model, by using continuous time Markov chain (CTMC), for a IaaS CDC, which (1) consists of active and standby physical machines (PMs), (2) allows PM migration among active and standby PM pools, (3) all jobs are homogeneous, and (4) a running job could continue its running by using idle active PMs when the PM working for this job fails. Although a monolithic CTMC model for IaaS Cloud performability analysis may face largeness and stiffness problems, it could be used to verify the scalable approximate model. We present the details of state transition rules of the proposed model and the formula for computing metrics, including the immediate service probability, the mean response time and so on. Numerical analysis and simulations are carried out to verify the accuracy of the proposed model.","PeriodicalId":203925,"journal":{"name":"2016 17th International Conference on Parallel and Distributed Computing, Applications and Technologies (PDCAT)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134243999","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dongsheng Yang, Yijie Wang, Yongmou Li, Xingkong Ma
Nowadays sequence data tends to be multi-dimensional sequence over data stream, it has a large state space and arrives at unprecedented speed. It is a big challenge to design a multi-dimensional sequence outlier detection method to meet the accurate and high speed requirements. The traditional methods can't handle multi-dimensional sequence effectively as they have poor abilities for multi-dimensional sequence modeling, and can't detect outlier timely as they have high computational complexity. In this paper we propose a variable Markovian based outlier detection method for multi-dimensional sequence over data stream, VMOD, which consists of two algorithms: mutual information based feature selection algorithm (MIFS), variable Markovian based sequential analysis algorithm (VMSA). It uses MIFS algorithm to reduce the state space and redundant features, and uses VMSA algorithm to accelerate the outlier detection. Through VMOD method, we can improve the detection rate and detection speed. The MIFS algorithm uses mutual information as similarity measures and adopt clustering based strategy to select features, it can improve the abilities for sequence modeling through reducing the state space and redundant features, consequently, to improve the detection rate. The VMSA algorithm use random sample and index structure to accelerate the variable Markovian model construction and reduce the model complexity, consequently, to quicken the outlier detection. The experiments show that VMOD can detect outlier effectively, and reduce the detection time by at least 50% compared with the traditional methods.
{"title":"A Variable Markovian Based Outlier Detection Method for Multi-Dimensional Sequence over Data Stream","authors":"Dongsheng Yang, Yijie Wang, Yongmou Li, Xingkong Ma","doi":"10.1109/PDCAT.2016.049","DOIUrl":"https://doi.org/10.1109/PDCAT.2016.049","url":null,"abstract":"Nowadays sequence data tends to be multi-dimensional sequence over data stream, it has a large state space and arrives at unprecedented speed. It is a big challenge to design a multi-dimensional sequence outlier detection method to meet the accurate and high speed requirements. The traditional methods can't handle multi-dimensional sequence effectively as they have poor abilities for multi-dimensional sequence modeling, and can't detect outlier timely as they have high computational complexity. In this paper we propose a variable Markovian based outlier detection method for multi-dimensional sequence over data stream, VMOD, which consists of two algorithms: mutual information based feature selection algorithm (MIFS), variable Markovian based sequential analysis algorithm (VMSA). It uses MIFS algorithm to reduce the state space and redundant features, and uses VMSA algorithm to accelerate the outlier detection. Through VMOD method, we can improve the detection rate and detection speed. The MIFS algorithm uses mutual information as similarity measures and adopt clustering based strategy to select features, it can improve the abilities for sequence modeling through reducing the state space and redundant features, consequently, to improve the detection rate. The VMSA algorithm use random sample and index structure to accelerate the variable Markovian model construction and reduce the model complexity, consequently, to quicken the outlier detection. The experiments show that VMOD can detect outlier effectively, and reduce the detection time by at least 50% compared with the traditional methods.","PeriodicalId":203925,"journal":{"name":"2016 17th International Conference on Parallel and Distributed Computing, Applications and Technologies (PDCAT)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129339838","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The high dimensionality of the current battlefield information increases the complexity of the information utilization, which leads to the deterioration of the battlefield information services. The effective reduction of the of battlefield information dimension by information feature selection is an important prerequisite for the effective development of battlefield information service. The traditional feature selection method is not applicable due to the absence of accurate labels of items in battlefield text information. An attribute reduction method based on set division is proposed and applied to the battlefield text feature selection. An improved document frequency (DF) method for text feature selection is used to filter noise words, then the text feature is selected by the attribute reduction based on set division. Experimental results demonstrate that the proposed feature selection algorithm is able to obtain a better feature subset of battlefield text information when compared with other existing feature selection algorithms.
{"title":"Text Feature Selection Method in Battlefield Information Service","authors":"Wang Kai, Gan Zhi-chun, L. Jingzhi, Cai Yan-jun","doi":"10.1109/PDCAT.2016.055","DOIUrl":"https://doi.org/10.1109/PDCAT.2016.055","url":null,"abstract":"The high dimensionality of the current battlefield information increases the complexity of the information utilization, which leads to the deterioration of the battlefield information services. The effective reduction of the of battlefield information dimension by information feature selection is an important prerequisite for the effective development of battlefield information service. The traditional feature selection method is not applicable due to the absence of accurate labels of items in battlefield text information. An attribute reduction method based on set division is proposed and applied to the battlefield text feature selection. An improved document frequency (DF) method for text feature selection is used to filter noise words, then the text feature is selected by the attribute reduction based on set division. Experimental results demonstrate that the proposed feature selection algorithm is able to obtain a better feature subset of battlefield text information when compared with other existing feature selection algorithms.","PeriodicalId":203925,"journal":{"name":"2016 17th International Conference on Parallel and Distributed Computing, Applications and Technologies (PDCAT)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121795105","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Peng Xiao, Na Liu, Yuanyuan Li, Ying Lu, Xiao-Jun Tang, Hai-Wen Wang, Ming-Xia Li
Traffic classification is becoming one of the major applications in the data center networks with a lot of cloud services. Recent works about software defined networking (SDN) have found new ways to manage data center networks. However, with the imbalance of the elephant and mice flows is sharpening, the accuracy and efficiency of traffic classification have become more and more important in SDN management. To address this issue, in this paper, we propose a traffic classification method that can deal with the traffic classification in SDN. Our method is based on spectral clustering and Software-Defined Networking (SDN). We propose a real-time flow extraction and representation method by scanning the flow tables in SDN controller. Then we cluster the flow data with spectral analysis. Extensive experiments on different settings have been performed, showing that our method is good at traffic classification with high detection rates and low overhead.
{"title":"A Traffic Classification Method with Spectral Clustering in SDN","authors":"Peng Xiao, Na Liu, Yuanyuan Li, Ying Lu, Xiao-Jun Tang, Hai-Wen Wang, Ming-Xia Li","doi":"10.1109/PDCAT.2016.089","DOIUrl":"https://doi.org/10.1109/PDCAT.2016.089","url":null,"abstract":"Traffic classification is becoming one of the major applications in the data center networks with a lot of cloud services. Recent works about software defined networking (SDN) have found new ways to manage data center networks. However, with the imbalance of the elephant and mice flows is sharpening, the accuracy and efficiency of traffic classification have become more and more important in SDN management. To address this issue, in this paper, we propose a traffic classification method that can deal with the traffic classification in SDN. Our method is based on spectral clustering and Software-Defined Networking (SDN). We propose a real-time flow extraction and representation method by scanning the flow tables in SDN controller. Then we cluster the flow data with spectral analysis. Extensive experiments on different settings have been performed, showing that our method is good at traffic classification with high detection rates and low overhead.","PeriodicalId":203925,"journal":{"name":"2016 17th International Conference on Parallel and Distributed Computing, Applications and Technologies (PDCAT)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123316145","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Data streams classification poses three major challenges, namely, infinite length, concept-drift, and featureevolution. The first two issues have been widely studied. However, most existing data stream classification techniques ignore the last one. DXMiner [17], the first model which addresses featureevolution by using the past labeled instances to select the top ranked features based on a scores computed by a formula. This semi-supervised feature selection method depends on the quality of the past classification and neglects the possible correlation among different features, thus unable to produce an optimal feature subset which deteriorates the accuracy of classification. Multi-Cluster Feature Selection (MCFS) [5] proposed for static data classification and clustering applies unsupervised feature selection to address the feature-evolution problem, but suffers from the high computational cost in feature selection. In this paper, we apply MCFS in the DXMiner framework to handle each window of data in a data stream for dynamic data stream-classification. With unsupervised feature selection, our method produces the optimal feature subset and hence improves DXMiner on the classification accuracy. We further improve the time complexity of the feature selection process in MCFS by using the locality sensitive hashing forest (LSH Forest) [4]. The empirical results indicate that our approach outperforms stateof-the-art streams classification techniques in classifying real-life data streams.
{"title":"Improved Data Streams Classification with Fast Unsupervised Feature Selection","authors":"Lulu Wang, Hong Shen","doi":"10.1109/PDCAT.2016.056","DOIUrl":"https://doi.org/10.1109/PDCAT.2016.056","url":null,"abstract":"Data streams classification poses three major challenges, namely, infinite length, concept-drift, and featureevolution. The first two issues have been widely studied. However, most existing data stream classification techniques ignore the last one. DXMiner [17], the first model which addresses featureevolution by using the past labeled instances to select the top ranked features based on a scores computed by a formula. This semi-supervised feature selection method depends on the quality of the past classification and neglects the possible correlation among different features, thus unable to produce an optimal feature subset which deteriorates the accuracy of classification. Multi-Cluster Feature Selection (MCFS) [5] proposed for static data classification and clustering applies unsupervised feature selection to address the feature-evolution problem, but suffers from the high computational cost in feature selection. In this paper, we apply MCFS in the DXMiner framework to handle each window of data in a data stream for dynamic data stream-classification. With unsupervised feature selection, our method produces the optimal feature subset and hence improves DXMiner on the classification accuracy. We further improve the time complexity of the feature selection process in MCFS by using the locality sensitive hashing forest (LSH Forest) [4]. The empirical results indicate that our approach outperforms stateof-the-art streams classification techniques in classifying real-life data streams.","PeriodicalId":203925,"journal":{"name":"2016 17th International Conference on Parallel and Distributed Computing, Applications and Technologies (PDCAT)","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127063633","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Word alignment is a basic task in natural language processing and it usually serves as the starting point when building a modern statistical machine translation system. However, the state-of-art parallel algorithm for word alignment is still time-consuming. In this work, we explore a parallel implementation of word alignment algorithm on Graphics Processor Unit (GPU), which has been widely available in the field of high performance computing. We use the Compute Unified Device Architecture (CUDA) programming model to re-implement a state-of-the-art word alignment algorithm, called IBM Expectation-Maximization (EM) algorithm. A Tesla K40M card with 2880 cores is used for experiments and execution times obtained with the proposed algorithm are compared with a sequential algorithm and a multi-threads algorithm on an IBM X3850 server, which has two Intel Xeon E7 CPUs (2.0GHz * 10 cores). The best experimental results show a 16.8-fold speedup compared to the multi-threads algorithm and a 234.7-fold speedup compared to the sequential algorithm.
{"title":"CUDA-Based Parallel Implementation of IBM Word Alignment Algorithm for Statistical Machine Translation","authors":"Siyuan Jing, Gaorong Yan, Xingyuan Chen, Peng Jin, Zhaoyi Guo","doi":"10.1109/PDCAT.2016.050","DOIUrl":"https://doi.org/10.1109/PDCAT.2016.050","url":null,"abstract":"Word alignment is a basic task in natural language processing and it usually serves as the starting point when building a modern statistical machine translation system. However, the state-of-art parallel algorithm for word alignment is still time-consuming. In this work, we explore a parallel implementation of word alignment algorithm on Graphics Processor Unit (GPU), which has been widely available in the field of high performance computing. We use the Compute Unified Device Architecture (CUDA) programming model to re-implement a state-of-the-art word alignment algorithm, called IBM Expectation-Maximization (EM) algorithm. A Tesla K40M card with 2880 cores is used for experiments and execution times obtained with the proposed algorithm are compared with a sequential algorithm and a multi-threads algorithm on an IBM X3850 server, which has two Intel Xeon E7 CPUs (2.0GHz * 10 cores). The best experimental results show a 16.8-fold speedup compared to the multi-threads algorithm and a 234.7-fold speedup compared to the sequential algorithm.","PeriodicalId":203925,"journal":{"name":"2016 17th International Conference on Parallel and Distributed Computing, Applications and Technologies (PDCAT)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114643249","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Traffic Matrix (TM) can contain information about irregular network topology structure and depict the traffic characteristics of global network. It is a critical parameter to network traffic engineering and attracts significant research interests. Diffusion Wavelet (DW) can perform an effective Multi-Resolution Analysis (MRA)on TM in both temporaland space domains because it intrinsically adapts to the underlying network structure. This paper shows how to apply DW to TM analysis and anomaly detection. By comparing with other anomaly detection methods, it is confirmed thatour method can detect anomaly effectively due to combining with the analysis results by DW.
{"title":"Diffusion Wavelet-Based Anomaly Detection in Networks","authors":"Hui Tian, Meimei Ding","doi":"10.1109/PDCAT.2016.087","DOIUrl":"https://doi.org/10.1109/PDCAT.2016.087","url":null,"abstract":"Traffic Matrix (TM) can contain information about irregular network topology structure and depict the traffic characteristics of global network. It is a critical parameter to network traffic engineering and attracts significant research interests. Diffusion Wavelet (DW) can perform an effective Multi-Resolution Analysis (MRA)on TM in both temporaland space domains because it intrinsically adapts to the underlying network structure. This paper shows how to apply DW to TM analysis and anomaly detection. By comparing with other anomaly detection methods, it is confirmed thatour method can detect anomaly effectively due to combining with the analysis results by DW.","PeriodicalId":203925,"journal":{"name":"2016 17th International Conference on Parallel and Distributed Computing, Applications and Technologies (PDCAT)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124231252","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With the development of the technology, Vehicular Ad-hoc Network is developing rapidly and continuously. And many new algorithms in VANET were put forward. For example, many people apply the Named Data Networks or Software Defined Networking6 to the VANET. However, few researches are looking into the analysis of the topology of the VANET, among most of which are based on simple situations and calculation of the track of vehicles. And therefore, these analysis are inadequate for analyzing the topology under complex situations. This paper is going to put forward a new method for analyzing the change of the topology of VANET, which is capable to quantitatively analyze the change under different and complex situations. In the real experiment, when compared with traditional analysis methods, this new method can analyze more precisely the impact of some possible wireless communication problem on the change of the topology, for example, the impact of hidden node on the vehicular wireless communication and whether the number or the density of vehicles makes the topology of VANET more stable. And this new method can provide a numerical result instead of purely qualitative analysis.
{"title":"Topology Analysis System for Vehicular Ad Hoc Network","authors":"Baihong Dong, Jian Deng, Weigang Wu, Tianyu Meng","doi":"10.1109/PDCAT.2016.070","DOIUrl":"https://doi.org/10.1109/PDCAT.2016.070","url":null,"abstract":"With the development of the technology, Vehicular Ad-hoc Network is developing rapidly and continuously. And many new algorithms in VANET were put forward. For example, many people apply the Named Data Networks or Software Defined Networking6 to the VANET. However, few researches are looking into the analysis of the topology of the VANET, among most of which are based on simple situations and calculation of the track of vehicles. And therefore, these analysis are inadequate for analyzing the topology under complex situations. This paper is going to put forward a new method for analyzing the change of the topology of VANET, which is capable to quantitatively analyze the change under different and complex situations. In the real experiment, when compared with traditional analysis methods, this new method can analyze more precisely the impact of some possible wireless communication problem on the change of the topology, for example, the impact of hidden node on the vehicular wireless communication and whether the number or the density of vehicles makes the topology of VANET more stable. And this new method can provide a numerical result instead of purely qualitative analysis.","PeriodicalId":203925,"journal":{"name":"2016 17th International Conference on Parallel and Distributed Computing, Applications and Technologies (PDCAT)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123068865","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Parallel many cores contribute to heterogeneous architectures and achieve high computation throughput. Working as coprocessors and connected to general-purpose CPUs via PCIe, those special-purpose cores usually work as float computing accelerators (ACC). The popular programming models typically offload the computing intensive parts to accelerator then aggregate results, which would result in a great amount of data transfer via PCIe. In this paper, we introduce an ACC-centered model to leverage the limited bandwidth of PCIe, increase performance, reduce idle time of ACC. In order to realize dada-near-computing, our ACC-centered model arms to program centered on ACC and the control intensive parts are offloaded to CPU. Both CPU and ACC are devoted to higher performance with their architect feature. Validation on the Tianhe-2 supercomputer shows that the implementation of ACC-centered LU competes with the highly optimized Intel MKL hybrid implementation and achieves about 5× speedup versus the CPU version.
{"title":"Accelerator-Centered Programming on Heterogeneous Systems","authors":"Cheng Chen, Yunfei Du, Canqun Yang","doi":"10.1109/PDCAT.2016.041","DOIUrl":"https://doi.org/10.1109/PDCAT.2016.041","url":null,"abstract":"Parallel many cores contribute to heterogeneous architectures and achieve high computation throughput. Working as coprocessors and connected to general-purpose CPUs via PCIe, those special-purpose cores usually work as float computing accelerators (ACC). The popular programming models typically offload the computing intensive parts to accelerator then aggregate results, which would result in a great amount of data transfer via PCIe. In this paper, we introduce an ACC-centered model to leverage the limited bandwidth of PCIe, increase performance, reduce idle time of ACC. In order to realize dada-near-computing, our ACC-centered model arms to program centered on ACC and the control intensive parts are offloaded to CPU. Both CPU and ACC are devoted to higher performance with their architect feature. Validation on the Tianhe-2 supercomputer shows that the implementation of ACC-centered LU competes with the highly optimized Intel MKL hybrid implementation and achieves about 5× speedup versus the CPU version.","PeriodicalId":203925,"journal":{"name":"2016 17th International Conference on Parallel and Distributed Computing, Applications and Technologies (PDCAT)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128909631","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Software-defined networking (SDN), as a new network paradigm, has the advantage of centralizing control and global visibility over a network. However, security issues remain a major concern and prevent SDN from being widely adopted. One of the challenges is the prevention of malicious OpenFlow application (OF app) access to the SDN controller as it opens a programmable northbound interface for third party applications. In this paper, we address app-to-control security issues with focus on five main attack vectors: unauthorized access, illegal function calling, malicious rules injection, resources exhausting and manin-the-middle attack. Based on the identified threat models, we develop a light-weight plug-in, which is called ControllerSEPA, by using RESTful API to defend SDN controller against malicious OF apps. Specifically, ControllerSEPA can provide the services including OF app-based AAA control (unlike OpenDaylight and ONOS which offer user-based or role-based AAA control), rule conflict resolution, OF app isolation, fine-grained access control and encryption. Furthermore, we study the feasibility of deploying ControllerSEPA on five open source SDN controllers: OpenDaylight, ONOS, Floodlight, Ryu and POX. Results show that the deployment operates with very low complexity, and most of time the modification of source codes is unnecessary. In our implementations, the repacked services in ControllerSEPA create negligible latency (0.1% to 0.3%) and can provide more rich services to OF apps.
{"title":"ControllerSEPA: A Security-Enhancing SDN Controller Plug-in for OpenFlow Applications","authors":"Yuchia Tseng, Zonghua Zhang, Farid Naït-Abdesselam","doi":"10.1109/PDCAT.2016.064","DOIUrl":"https://doi.org/10.1109/PDCAT.2016.064","url":null,"abstract":"Software-defined networking (SDN), as a new network paradigm, has the advantage of centralizing control and global visibility over a network. However, security issues remain a major concern and prevent SDN from being widely adopted. One of the challenges is the prevention of malicious OpenFlow application (OF app) access to the SDN controller as it opens a programmable northbound interface for third party applications. In this paper, we address app-to-control security issues with focus on five main attack vectors: unauthorized access, illegal function calling, malicious rules injection, resources exhausting and manin-the-middle attack. Based on the identified threat models, we develop a light-weight plug-in, which is called ControllerSEPA, by using RESTful API to defend SDN controller against malicious OF apps. Specifically, ControllerSEPA can provide the services including OF app-based AAA control (unlike OpenDaylight and ONOS which offer user-based or role-based AAA control), rule conflict resolution, OF app isolation, fine-grained access control and encryption. Furthermore, we study the feasibility of deploying ControllerSEPA on five open source SDN controllers: OpenDaylight, ONOS, Floodlight, Ryu and POX. Results show that the deployment operates with very low complexity, and most of time the modification of source codes is unnecessary. In our implementations, the repacked services in ControllerSEPA create negligible latency (0.1% to 0.3%) and can provide more rich services to OF apps.","PeriodicalId":203925,"journal":{"name":"2016 17th International Conference on Parallel and Distributed Computing, Applications and Technologies (PDCAT)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116493513","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：