一种改进的核心无状态公平带宽分配机制

CSFQ(Core Stateless Fair Queue)算法在无状态网实现了如同有状态网那样好的公平带宽分配，但它的丢包算法是针对UDP流等非响应流导出的，不适用于TCP流．本文针对TCP流的特点提出了一种改进的CSFQ算法．算法主要改进有两点：(1)将缓存队列长度变化与丢包概率关联起来，用一种类似于RED(Random Early Drop)的缓存策略解决了缓存频繁溢出导致的一些问题，如公平共享速率的收敛问题；(2)对TCP流使用了与UDP流不同的丢包策略，彻底解决TCP流与UDP流的带宽分配公平性．仿真结果显示，当TCP流与UDP流共享拥塞链路时，在带宽分配的公平性方面，改进算法较原算法有了很大的性能提高．     

基于CSFQ的路由器带宽公平性算法的改进研究

针对核心无状态公平队列调度(CSFQ)的丢包算法不适用于TCP流的问题,提出了一种改进算法my-CS-FQ,能够根据网络状况动态地调整TCP流和UDP流各自的丢包行为,解决TCP流与UDP流的带宽分配公平性.大量仿真实验结果表明my-CSFQ算法能够提高TCP流同UDP流竞争网络资源的能力,更有效地利用链路带宽,提高网...     

基于SDN的数据中心网络中考虑等待时间的流调度策略研究

随着云计算、大数据、工业互联网的发展,越来越多的应用服务被部署到数据中心,其中一些数据密集型应用程序要求流在尽可能短的时间内完成.提出了一种考虑等待时间的流调度策略(flow priority scheduling algorithm considering wait time,FPWT),该策略在划分流的优先级时不仅...     

光突发交换网络中TCP短流的分析模型

分析了光突发交换(OBS)网络中突发传送方式对TCP短流的影响,以平均发送完成时间为主要性能指标,提出了一种OBS网络中TCP短流的分析模型,导出了一个平均发送完成时间与TCP短流长度、突发丢失率、突发组装时间以及往返时间等之间的递推表达式.仿真验证了分析的正确性,结果表明,当突发丢失率小于0.01且突发组装时间与往返时间之比小于0.1时,突发传送方式对TCP短流的性能没有影响.     

云存储数据缩减技术研究

基于云计算应用中的云存储技术,使数据存储变得安全可靠和易管理。在云存储技术数据存储的过程中,不仅考虑数据读写的速度,还得处理数据存储效率,以便满足当前海量信息存储的需求。云存储中的数据缩减技术可以缩减数据信息量,提高存储的效率,满足数据存储急剧的要求。通过对几种数据缩减技术进行比较研究分析,探讨了对数据处理后存储的效率以及每项技术发展状况,为用户选择云存储数据缩减技术提供有力的参考。     

A partition checkpoint strategy based on data segment priority

A partition checkpoint strategy based on data segment priority is presented to meet the timing constraints of the data and the transaction in embedded real-time main memory database systems(ERTMMDBS) as well as to reduce the number of the transactions missing their deadlines and the recovery time.The partition checkpoint strategy takes into account the characteristics of the data and the transactions associated with it;moreover,it partitions the database according to the data segment priority and sets the corresponding checkpoint frequency to each partition for independent checkpoint operation.The simulation results show that the partition checkpoint strategy decreases the ratio of trans-actions missing their deadlines.     

基于目标检测的激光雷达道路边缘检测算法

道路边缘检测是自动驾驶车辆环境感知的重要组成部分,有效地从点云数据中提取道路边缘信息,有利于进行目标检测以及可行驶区域检测。针对点云道路边缘检测问题,提出了一种考虑车辆等道路参与者对道路边缘检测带来干扰的解决方案。首先,采用地面点云分割算法,将原始点云分割成地面点云和非地面点云;其次,根据车辆等道路参与者的固有特性,采用点云聚类算法对点云进行聚类,并将符合车辆等道路参与者特性的非地面点云进行滤除;再次,根据道路边缘点云在二维平面内,能够有效地遮挡激光发射中心点与非道路边缘点之间的连线,从而提取道路边缘点云;最后,采用随机抽样一致性(random sample consensus, RANSAC)算法对道路边缘点云进行多项式拟合,并使用扩展卡尔曼滤波器对道路边缘进行跟踪。实验结果表明,所提点云道路边缘检测算法能够消除车辆等道路参与则对点云道路边缘检测的影响,且算法满足实车实时性和鲁棒性要求。     

基于历史特征的FAST TCP公平性改进算法

新建和已存在的旧FAST TCP连接估测的传播延时不同,它们无法公平地分配瓶颈链路的带宽。为了解决新旧FAST TCP连接的公平性,提出了一种改进估测FAST TCP传播延时的算法。在各连接无法通信的情况下,该改进算法利用新连接建立时,旧连接会表现RTT逐步增大和拥塞窗口逐步减小的特征。当旧连接发现这一特征出现时,就同步计算新的“传播延时”,从而使新旧连接具有相同的“排队延时”,公平地分配瓶颈链路带宽。NS-2仿真实例验证了该改进算法的有效性。     

用MAC层突发传输提高无线局域网TCP公平性

在无线局域网(W LAN)系统中,IEEE 802.11协议在媒介访问控制(M AC)层对各个站点提供公平的信道接入机会。然而由于中心站点的业务量远远大于其他移动站点,M AC层的公平性恶化了传输控制协议(TCP)上、下行业务的公平性。该文以M AC层的观点分析了造成TCP业务不公平性的原因以及现有改进算法中存在的问题。提出了一种利用M AC层突发传输提高中心站点优先级的算法,并采用成批服务M/M/1/B排队模型分析了该算法的性能。仿真结果显示,该算法能够很好地保证TCP上、下行业务流的公平性,大量节省缓冲区容量,而且具有实现简单的优点。     

An adaptive mechanism to guarantee the bandwidth fairness of TCP flows

End-to-end TCP (transmission control protocol) congestion control can cause unfairness among multiple TCP connections with different RTT (Round Trip Time). The throughput of TCP connection is inversely proportional to its RTT. To resolve this problem, researchers have proposed many methods. The existing proposals for RTT-aware conditioner work well when congestion level is low. However, they over-protect long RTT flows and starve short RTT flows when congestion level is high. Due to this reason, an improved method based on adaptive thought is proposed. According to the congestion level of networks, the mechanism can adaptively adjust the degree of the protection to long RTT flows. Extensive simulation experiments showed that the proposed mechanism can guarantee the bandwidth fairness of TCP flows effectively and outperforms the existing methods.     

On Peer-Assisted Data Dissemination in Data Center Networks: Analysis and Implementation简

Data Center Networks （DCNs） are the fundamental infrastructure for cloud computing. Driven by the massive parallel computing tasks in cloud computing, one-to-many data dissemination becomes one of the most important traffic patterns in DCNs. Many architectures and protocols are proposed to meet this demand. However, these proposals either require complicated configurations on switches and servers, or cannot deliver an optimal performance. In this paper, we propose the peer-assisted data dissemination for DCNs. This approach utilizes the rich physical connections with high bandwidths and mutli-path connections, to facilitate efficient one-to-many data dissemination. We prove that an optimal P2P data dissemination schedule exists for FatTree, a specially- designed DCN architecture. We then present a theoretical analysis of this algorithm in the general multi-rooted tree topology, a widely-used DCN architecture. Additionally, we explore the performance of an intuitive line structure for data dissemination. Our analysis and experimental results prove that this simple structure is able to produce a comparable performance to the optimal algorithm. Since DCN applications heavily rely on virtualization to achieve optimal resource sharing, we present a general implementation method for the proposed algorithms, which aims to mitigate the impact of the potentially-high churn rate of the virtual machines.     

2003—2020年新疆北部积雪因子时空变化分析

积雪作为冰冻圈的活跃因子之一,对气候环境的敏感性使其能够快速反映出与气温、降水等气候因素的关系变化,并影响着全球水文变化。对新疆北部地区2003年6月—2021年6月的中分辨率成像光谱仪(moderate resolution imaging spectrometer, MODIS)逐日积雪数据进行去云处理,并基于像元计算了积雪覆盖比例(snow cover percentage, SCP)、积雪覆盖日数(snow cover days, SCD)、积雪开始时间(snow onset day, SOD)和积雪结束时间(snow end day, SED)。实验结果表明：积雪产品相对于气象站点数据的总精度达到91.47%,有利于提高对积雪因子的时空变化分析。新疆北部地区SCD空间分布差异较大,SCD随海拔的升高而增加,SOD随海拔升高而提前,主要在11月和12月出现;SED随海拔升高而推迟,主要在2月和3月出现;夏季平均SCD最少,主要分布于天山中部以及阿尔泰山北部区域,约占3.35%,冬季平均SCD最为明显,大于60 d的区域占46.3%;而SCP在1月达到最大,7月和8月最小。趋势变...     

Discovery of molecular hydrogen in a high-velocity cloud of the Galactic halo

The Milky Way's halo contains clouds of neutral hydrogen with high radial velocities which do not follow the general rotational motion of the Galaxy. Few distances to these high-velocity clouds are known, so even gross properties such as total mass are hard to determine. As a consequence, there is no generally accepted theory regarding their origin. One idea is that they result from gas that has cooled after being ejected from the Galaxy through fountain-like flows powered by supernovae; another is that they are composed of gas, poor in heavy elements, which is falling onto the disk of the Milky Way from intergalactic space. The presence of molecular hydrogen, whose formation generally requires the presence of dust (and therefore gas, enriched in heavy elements), could help to distinguish between these possibilities. Here we report the discovery of molecular hydrogen absorption in a high-velocity cloud along the line of sight to the Large Magellanic Cloud. We also derive for the same cloud an iron abundance which is half of the solar value. From these data, we conclude that gas in this cloud originated in the disk of the Milky Way.     

A cost-effective cloud leasing policy for live streaming applications

The dynamics of the globalized multimedia sources and request demands, which requires high computations and bandwidths, makes the IT infrastructure a challenge for live streaming applications. Migrating the system to a geo-distributed cloud and leasing servers is an ideal alternative for supporting large-scale live streaming applications with dynamic contents and demands. The new challenge of multimedia live streaming applications in a geo-distributed cloud is how to efficiently arrange and migrate services among different cloud sites to guarantee the distribute users’ experience at modest costs. This paper first investigates cloud leasing policies for live streaming applications and finds that there is no detailed algorithm to help live streaming applications arrange and migrate services among different cloud sites. Then, we present a quality of service (QoS) guarantee cost-effective cloud leasing policy for live streaming applications. Meanwhile, we design a genetic algorithm (GA) to deal with the leasing policy among cloud sites of diverse lease prices. Experimental results confirm the effectiveness of the proposed model and the efficiency of the involved GA.     

基于贝叶斯信任模型的资源调度策略

在科技迅速发展的今天,云平台正在迅速成为科学应用的重要平台.由于云平台下的节点资源数量非常多,显然他们有很多是不可靠的,这些节点会对EXE的执行和调度任务产生很大的影响.在本文中,受贝叶斯认知模型的启发和社会学的信任关系模型的引导,笔者首先提出了一种新的基于贝叶斯方法的认知信任模型,然后,将这种模型应用到资源调度系统中.理论分析和仿真实验证明,笔者提出的方法能有效的满足云计算对节点资源的信任要求,并且牺牲较少的时间成本,确保在一个相对安全的节点资源池中执行云计算任务.     

TCP连接传输速率限制因素的测量和诊断

TCP连接的传输速率是很重要的一个网络性能指标.由于很多应用的性能都取决于它,我们往往需要找出TCP连接传输速率的瓶颈并进而提高.然而,TCP连接传输速率受到多种因素如丢包、瓶颈带宽以及发送接收端缓存的限制,分析起来比较复杂.本文描述了我们设计的一个工具,它通过分析TCP数据trace文件来确定连接发送窗口的受限因素并测量连接的RTT,使我们很容易确定连接传输速率的受限因素.实际网络测量表明该工具能够分辨多种传输速率限制因素.     

A study on TCP performance of crowdsourced live streaming

The prevalence of smart phone and improvement of wireless net promote the usage of crowdsourced live streaming,where individual users act as live streaming sources to broadcast themselves online.Characterizing the performance and identifying its bottleneck in such systems can shed light on the system design and performance optimization.TCP performance of a commercial crowdsourced live streaming system is examined by analyzing packet-level traces collected at streaming servers.TCP stalls that heavily hurt the QoE of user have been identified.In particular,the TCP stalls account for as much as 31.6％ of the flow completion time for upload flows and result in abandonment of upload on the corresponding channels.Stalls caused by timeout retransmissions are further dissected and timeout retransmission characteristics are revealed to be dependent on the video encoding methods.These findings provide new insights in crowdsourced live streaming systems and can guide designers to improve the TCP efficiency.     

Cluster as a Service: A Resource Sharing Approach for Private Cloud

With the rapid development of cloud computing and big data processing, an increasing number of application frameworks are being considered to run in a "cloud way". This development brings about several challenges to the enterprise private cloud computing platform, e.g., being able to run most existing heterogeneous applications, providing scalability and elasticity support for newly emerged frameworks, and most importantly,sharing cluster resources effectively. In this paper, we propose a new service model, namely, Cluster as a Service(Claa S), which is suitable for medium- and small-sized data centers to solve these problems in a relatively easy and general way. The idea behind this model is virtualizing the cluster environment for distributed application frameworks. Most applications can directly run in the virtual cluster environment without any modification, which is a great advantage. Based on lightweight containers, we implement a real system of Claa S named Docklet to prove the feasibility of this service model. Meanwhile, we preliminarily design the definition of applications to make them easy to deploy. Finally, we present several examples and evaluate the entire system.     

Efficient Multi-Tenant Virtual Machine Allocation in Cloud Data Centers

Virtual Machine(VM) allocation for multiple tenants is an important and challenging problem to provide efficient infrastructure services in cloud data centers. Tenants run applications on their allocated VMs, and the network distance between a tenant's VMs may considerably impact the tenant's Quality of Service(Qo S). In this study, we define and formulate the multi-tenant VM allocation problem in cloud data centers, considering the VM requirements of different tenants, and introducing the allocation goal of minimizing the sum of the VMs' network diameters of all tenants. Then, we propose a Layered Progressive resource allocation algorithm for multi-tenant cloud data centers based on the Multiple Knapsack Problem(LP-MKP). The LP-MKP algorithm uses a multi-stage layered progressive method for multi-tenant VM allocation and efficiently handles unprocessed tenants at each stage. This reduces resource fragmentation in cloud data centers, decreases the differences in the Qo S among tenants, and improves tenants' overall Qo S in cloud data centers. We perform experiments to evaluate the LP-MKP algorithm and demonstrate that it can provide significant gains over other allocation algorithms.     

Collaborative Network Security in Multi-Tenant Data Center for Cloud Computing简

A data center is an infrastructure that supports Internet service. Cloud comput the face of the Internet service infrastructure, enabling even small organizations to quickly ng is rapidly changing build Web and mobile applications for millions of users by taking advantage of the scale and flexibility of shared physical infrastructures provided by cloud computing. In this scenario, multiple tenants save their data and applications in shared data centers, blurring the network boundaries between each tenant in the cloud. In addition, different tenants have different security requirements, while different security policies are necessary for different tenants. Network virtualization is used to meet a diverse set of tenant-specific requirements with the underlying physical network enabling multi-tenant datacenters to automatically address a large and diverse set of tenants requirements. In this paper, we propose the system implementation of vCNSMS, a collaborative network security prototype system used n a multi-tenant data center. We demonstrate vCNSMS with a centralized collaborative scheme and deep packet nspection with an open source UTM system. A security level based protection policy is proposed for simplifying the security rule management for vCNSMS. Different security levels have different packet inspection schemes and are enforced with different security plugins. A smart packet verdict scheme is also integrated into vCNSMS for ntelligence flow processing to protect from possible network attacks inside a data center network