In order to improve the precision of super point detection and control measurement resource consumption, this paper proposes a super point detection method based on sampling and data streaming algorithms （SDSD）, and...In order to improve the precision of super point detection and control measurement resource consumption, this paper proposes a super point detection method based on sampling and data streaming algorithms （SDSD）, and proves that only sources or destinations with a lot of flows can be sampled probabilistically using the SDSD algorithm. The SDSD algorithm uses both the IP table and the flow bloom filter （BF） data structures to maintain the IP and flow information. The IP table is used to judge whether an IP address has been recorded. If the IP exists, then all its subsequent flows will be recorded into the flow BF; otherwise, the IP flow is sampled. This paper also analyzes the accuracy and memory requirements of the SDSD algorithm , and tests them using the CERNET trace. The theoretical analysis and experimental tests demonstrate that the most relative errors of the super points estimated by the SDSD algorithm are less than 5%, whereas the results of other algorithms are about 10%. Because of the BF structure, the SDSD algorithm is also better than previous algorithms in terms of memory consumption.展开更多
基于分治和按需传输思想的分块传输技术是解决三维全息视频流传输的有效手段.然而,现有的分块方案要么缺乏自适应机制,要么不适用于移动实时通信场景.为此,本文提出了VVSTiler(Volumetric Video Streaming Tiling selector),一种面向全...基于分治和按需传输思想的分块传输技术是解决三维全息视频流传输的有效手段.然而,现有的分块方案要么缺乏自适应机制,要么不适用于移动实时通信场景.为此,本文提出了VVSTiler(Volumetric Video Streaming Tiling selector),一种面向全息视频通信的自适应分块传输方法,能够在动态且有限的计算和带宽资源下最大化视频的观感质量.具体而言,本文对不同粒度的分块方案带来的影响进行了初步研究,发现细粒度的分块方案可提高动态网络资源的利用率,粗粒度的分块方案可保证视频编解码效率和鲁棒性.基于此,本文构建了考虑预测视口、可用计算资源以及网络带宽等上下文信息的视频观感质量优化问题,并设计了一个高效的求解方案以支持在线的分块粒度决策.本文在8iVFB(8i Voxelized Full Bodies)标准数据集上将VVSTiler与当前主流的分块传输方法进行了比较.实验结果表明,VVSTiler在有偏差的视口预测情况下实现了高达60.4%的视频观感质量提升,在较准确的视口预测情况下平均每帧视频节省了27%的带宽资源.展开更多
In the context of Industry 4.0,a paradigm shift from traditional industrial manipulators to Collaborative Robots(CRs)is ongoing,with the latter serving ever more closely humans as auxiliary tools in many production pr...In the context of Industry 4.0,a paradigm shift from traditional industrial manipulators to Collaborative Robots(CRs)is ongoing,with the latter serving ever more closely humans as auxiliary tools in many production processes.In this scenario,continuous technological advancements offer new opportunities for further innovating robotics and other areas of next-generation industry.For example,6G could play a prominent role due to its human-centric view of the industrial domains.In particular,its expected dependability features will pave the way for new applications exploiting highly effective Digital Twin(DT)-and eXtended Reality(XR)-based telepresence.In this work,a novel application for the above technologies allowing two distant users to collaborate in the programming of a CR is proposed.The approach encompasses demanding data flows(e.g.,point cloud-based streaming of collaborating users and robotic environment),with network latency and bandwidth constraints.Results obtained by analyzing this approach from the viewpoint of network requirements in a setup designed to emulate 6G connectivity indicate that the expected performance of forthcoming mobile networks will make it fully feasible in principle.展开更多
The paper aims to make the abstract and even profound writing techniques of stream -of -consciousness concrete and easy -understanding as well as to demonstrate its functions in novel writing with illustration of the ...The paper aims to make the abstract and even profound writing techniques of stream -of -consciousness concrete and easy -understanding as well as to demonstrate its functions in novel writing with illustration of the famous work To the Lighthouse by Virginia Woolf who successfully employs this writing technique. The writing techniques involved are indirect interior monologue montage, free association and multiple -point -of view. The application of these approaches attributes greatly to better exposing t...展开更多
概念漂移是动态流数据挖掘中一类常见的问题,但混杂噪声或训练样本规模过小而产生的伪概念漂移会引起与真实概念漂移相似的结果,即模型在线测试性能的不稳定波动,导致二者容易混淆,发生概念漂移的误报.针对流数据中真伪概念漂移的混淆问...概念漂移是动态流数据挖掘中一类常见的问题,但混杂噪声或训练样本规模过小而产生的伪概念漂移会引起与真实概念漂移相似的结果,即模型在线测试性能的不稳定波动,导致二者容易混淆,发生概念漂移的误报.针对流数据中真伪概念漂移的混淆问题,提出一种基于在线性能测试的概念漂移检测方法(concept drift detection method based on online performance test,简称CDPT).该方法将最新获得的数据集进行均匀分组,在每组子数据集上分别进行在线学习,同时记录每组子数据集训练测试得到的分类精度向量,并计算相邻学习时间单元之间的精度落差,依据测试精度下降阈值得到有效波动位点.然后采用交叉检验的方式整合不同分组中的有效波动位点,以消除流数据在线学习过程中由于训练样本过小导致模型不稳定造成的检测干扰,根据精度波动一致性得到一致波动位点.最后,通过跟踪在线学习分类准确率,得到一致波动位点邻域参照点的测试精度变化,比较一致波动位点邻域参照点对应的模型测试精度下降幅度及收敛情况,以有效检测一致波动位点当中真实的概念漂移位点.实验结果表明,该方法能够有效辨识流数据在线学习过程中发生的真实概念漂移,并能有效避免训练样本过小或者流数据中噪声对检测结果的负面影响,同时提高模型的泛化性能.展开更多
超点检测对于网络安全、网络管理等应用具有重要意义.由于存在着高速网络环境下海量网络流量与有限系统资源之间的矛盾,在线准确地监测网络流量是一个极大的挑战.随着多核处理器的发展,多核处理器的并行性成为算法性能提高的一种有效途...超点检测对于网络安全、网络管理等应用具有重要意义.由于存在着高速网络环境下海量网络流量与有限系统资源之间的矛盾,在线准确地监测网络流量是一个极大的挑战.随着多核处理器的发展,多核处理器的并行性成为算法性能提高的一种有效途径.目前,针对基于流抽样的超点检测方法存在计算负荷重、检测精度低、实时性差等问题,提出了一种并行数据流方法(parallel data streaming,简称PDS).该方法构造并行的可逆Sketch数据结构,建立紧凑的节点链接度概要,在未存储节点地址信息的情况下,通过简单地计算重构超点的地址,获得了良好的效率和精度.实验结果表明:与CSE(compact spread estimator),JM(joint data streaming and sampling method)方法相比,该方法具有较好的性能,能够满足高速网络流量监测的应用需求.展开更多
基金The National Basic Research Program of China(973Program)(No.2009CB320505)the Natural Science Foundation of Jiangsu Province(No. BK2008288)+1 种基金the Excellent Young Teachers Program of Southeast University(No.4009001018)the Open Research Program of Key Laboratory of Computer Network of Guangdong Province (No. CCNL200706)
文摘In order to improve the precision of super point detection and control measurement resource consumption, this paper proposes a super point detection method based on sampling and data streaming algorithms （SDSD）, and proves that only sources or destinations with a lot of flows can be sampled probabilistically using the SDSD algorithm. The SDSD algorithm uses both the IP table and the flow bloom filter （BF） data structures to maintain the IP and flow information. The IP table is used to judge whether an IP address has been recorded. If the IP exists, then all its subsequent flows will be recorded into the flow BF; otherwise, the IP flow is sampled. This paper also analyzes the accuracy and memory requirements of the SDSD algorithm , and tests them using the CERNET trace. The theoretical analysis and experimental tests demonstrate that the most relative errors of the super points estimated by the SDSD algorithm are less than 5%, whereas the results of other algorithms are about 10%. Because of the BF structure, the SDSD algorithm is also better than previous algorithms in terms of memory consumption.
文摘基于分治和按需传输思想的分块传输技术是解决三维全息视频流传输的有效手段.然而,现有的分块方案要么缺乏自适应机制,要么不适用于移动实时通信场景.为此,本文提出了VVSTiler(Volumetric Video Streaming Tiling selector),一种面向全息视频通信的自适应分块传输方法,能够在动态且有限的计算和带宽资源下最大化视频的观感质量.具体而言,本文对不同粒度的分块方案带来的影响进行了初步研究,发现细粒度的分块方案可提高动态网络资源的利用率,粗粒度的分块方案可保证视频编解码效率和鲁棒性.基于此,本文构建了考虑预测视口、可用计算资源以及网络带宽等上下文信息的视频观感质量优化问题,并设计了一个高效的求解方案以支持在线的分块粒度决策.本文在8iVFB(8i Voxelized Full Bodies)标准数据集上将VVSTiler与当前主流的分块传输方法进行了比较.实验结果表明,VVSTiler在有偏差的视口预测情况下实现了高达60.4%的视频观感质量提升,在较准确的视口预测情况下平均每帧视频节省了27%的带宽资源.
基金funded by the European Commission through the H2020 project Hexa-X(Grant Agreement no.101015956).
文摘In the context of Industry 4.0,a paradigm shift from traditional industrial manipulators to Collaborative Robots(CRs)is ongoing,with the latter serving ever more closely humans as auxiliary tools in many production processes.In this scenario,continuous technological advancements offer new opportunities for further innovating robotics and other areas of next-generation industry.For example,6G could play a prominent role due to its human-centric view of the industrial domains.In particular,its expected dependability features will pave the way for new applications exploiting highly effective Digital Twin(DT)-and eXtended Reality(XR)-based telepresence.In this work,a novel application for the above technologies allowing two distant users to collaborate in the programming of a CR is proposed.The approach encompasses demanding data flows(e.g.,point cloud-based streaming of collaborating users and robotic environment),with network latency and bandwidth constraints.Results obtained by analyzing this approach from the viewpoint of network requirements in a setup designed to emulate 6G connectivity indicate that the expected performance of forthcoming mobile networks will make it fully feasible in principle.
文摘The paper aims to make the abstract and even profound writing techniques of stream -of -consciousness concrete and easy -understanding as well as to demonstrate its functions in novel writing with illustration of the famous work To the Lighthouse by Virginia Woolf who successfully employs this writing technique. The writing techniques involved are indirect interior monologue montage, free association and multiple -point -of view. The application of these approaches attributes greatly to better exposing t...
文摘概念漂移是动态流数据挖掘中一类常见的问题,但混杂噪声或训练样本规模过小而产生的伪概念漂移会引起与真实概念漂移相似的结果,即模型在线测试性能的不稳定波动,导致二者容易混淆,发生概念漂移的误报.针对流数据中真伪概念漂移的混淆问题,提出一种基于在线性能测试的概念漂移检测方法(concept drift detection method based on online performance test,简称CDPT).该方法将最新获得的数据集进行均匀分组,在每组子数据集上分别进行在线学习,同时记录每组子数据集训练测试得到的分类精度向量,并计算相邻学习时间单元之间的精度落差,依据测试精度下降阈值得到有效波动位点.然后采用交叉检验的方式整合不同分组中的有效波动位点,以消除流数据在线学习过程中由于训练样本过小导致模型不稳定造成的检测干扰,根据精度波动一致性得到一致波动位点.最后,通过跟踪在线学习分类准确率,得到一致波动位点邻域参照点的测试精度变化,比较一致波动位点邻域参照点对应的模型测试精度下降幅度及收敛情况,以有效检测一致波动位点当中真实的概念漂移位点.实验结果表明,该方法能够有效辨识流数据在线学习过程中发生的真实概念漂移,并能有效避免训练样本过小或者流数据中噪声对检测结果的负面影响,同时提高模型的泛化性能.
文摘超点检测对于网络安全、网络管理等应用具有重要意义.由于存在着高速网络环境下海量网络流量与有限系统资源之间的矛盾,在线准确地监测网络流量是一个极大的挑战.随着多核处理器的发展,多核处理器的并行性成为算法性能提高的一种有效途径.目前,针对基于流抽样的超点检测方法存在计算负荷重、检测精度低、实时性差等问题,提出了一种并行数据流方法(parallel data streaming,简称PDS).该方法构造并行的可逆Sketch数据结构,建立紧凑的节点链接度概要,在未存储节点地址信息的情况下,通过简单地计算重构超点的地址,获得了良好的效率和精度.实验结果表明:与CSE(compact spread estimator),JM(joint data streaming and sampling method)方法相比,该方法具有较好的性能,能够满足高速网络流量监测的应用需求.
