A coordinated data collection approach: design, evaluation, and comparison

We consider the problem of collecting a large amount of data from several different hosts to a single destination in a wide-area network. This problem is important since improvements in data collection times in many applications such as wide-area upload applications, high-performance computing applications, and data mining applications are crucial to performance of those applications. Often, due to congestion conditions, the paths chosen by the network may have poor throughput. By choosing an alternate route at the application level, we may be able to obtain substantially faster completion time. This data collection problem is a nontrivial one because the issue is not only to avoid congested link(s), but to devise a coordinated transfer schedule which would afford maximum possible utilization of available network resources. Our approach for computing coordinated data collection schedules makes no assumptions about knowledge of the topology of the network or the capacity available on individual links of the network. This approach provides significant performance improvements under various degrees and types of network congestions. To show this, we give a comprehensive comparison study of the various approaches to the data collection problem which considers performance, robustness, and adaptation characteristics of the different data collection methods. The adaptation to network conditions characteristics are important as the above applications are long lasting, i.e., it is likely changes in network conditions will occur during the data transfer process. In general, our approach can be used for solving arbitrary data movement problems over the Internet. We use the Bistro platform to illustrate one application of our techniques.     

基于NetFlow流记录的高速应用流量分类方法

针对目前应用流量分类算法效率不高的现状,提出一种以NetFlow统计的IP流记录信息作为输入的高速应用流量分类(FATC,fast application-level traffic classification)算法。该算法采用基于简单相关系数的测度选择算法衡量测度变量间的相关关系,删除对分类无用或相互冗余的测度,而后使用基于Bayes判别法的分类算法将网络流量分至误判损失最小的应用类别中。理论分析及实验表明,FATC算法在具有超过95%的分类准确率基础上,极大降低了当前应用流量分类方法在训练和分类过程的时空复杂度,满足实时准确分类当前10Gbit/s主干信道网络流量的需求。     

Performance Evaluation of Video,Voice and Web Traffic Over Heterogeneous Access Networks

By simulation using NS-3 we evaluated the performance of voice, video and web traffic sharing a wireless access network connected to a wired core. We compared the performance in terms of end-to-end delay, end-to-end delay variation, average throughput and loss percentage. For the wireless access network, we considered cases when it consisted of a single technology type, e.g., WiFi (IEEE 802.11), WiMAX (IEEE 802.16) and LTE, and when it was heterogeneous, i.e., when the three technologies coexisted and simultaneously shared the same IP core. We attempted to ascertain the impact of this type of heterogeneity on end-to-end performance. It was found that this heterogeneity in the wireless access portion of the network can improve, degrade or have no impact on application performance depending on the network conditions and the application itself. Some key research challenges in Fifth Generation wireless communications are heterogeneous Cloud Radio Access Networks (HC-RANS), backward compatiblity with 4G/3G networks and providing low-latency and QoE. To achieve end-to-end QoS guarantees in such settings the interface with the core must also be addressed, especially when backward compatibility is to be assured. This simulation study attempts to highlight the impact of this type of heterogeneity on network performance.     

Capacity estimation and TCP performance enhancement over mobile WiMAX networks - [WiMAX update]

The mobile WiMAX system is based on IEEE 802.16e, which defines radio interface supporting several classes of Internet Protocol applications and services. While the mobile WiMAX system is being deployed, IEEE 802.16m TG is developing an amendment to the IEEE 802.16e to greatly improve the system performance, and it is focusing not only on the PHY and MAC performance but also on a level of end-to-end performance improvement that includes the scope of the network and application to embrace the strong market request and interest. To evaluate the mobile WiMAX system capacity and performance, all the aspects of performance evaluation ? from air link to application ? are required. For the network and application-level capacity and performance analysis, we first provide an overview of mobile WiMAX systems, especially of the OFDMA/TDD systems of IEEE 802.16e and then describe subscriber and application profiles that include traffic-mix ratio, data-session attempts for applications, diurnal-application traffic distribution, and the application-traffic model. Afterward, the simulation results of network- traffic characteristics and demand estimation are provided. Finally, in the last section, we provide simulation results of end-to-end application performance evaluation using the examples of VoIP and a TCP/IP performance-enhancement method that can be implemented in the mobile WiMAX MAC or MAC/IP cross layer.     

Contradictory relationship between Hurst parameter and queueing performance (extended version)

Long Range Dependent (LRD) network traffic does not behave like the traffic generated by the Poisson model or other Markovian models. From the network performance point of view, the main difference is that LRD traffic increases queueing delays due to its burstiness over many time scales. LRD behavior has been observed in different types and sizes of networks, for different applications (e.g., WWW) and different traffic aggregations. Since LRD behaviour is not rare nor isolated, accurate characterization of LRD traffic is very important in order to predict performance and to allocate network resources. The Hurst parameter is commonly used to quantify the degree of LRD and the burstiness of the traffic. In this paper we investigate the validity and effectiveness of the Hurst parameter. To this end, we analyze the UCLA Computer Science Department network traffic traces and compute their Hurst parameters. Queueing simulation is used to study the impact of LRD and to determine if the Hurst parameter accurately describes such LRD. Our results show that the Hurst parameter is not by itself an accurate predictor of the queueing performance for a given LRD traffic trace.     

Analysis of the effects of SS7 message discard schemes on callcompletion rates during overload

The author considers the performance of a Signaling System Number 7 network when the routing processors, as opposed to transmission facilities, of signaling transfer points are overloaded. The choice of overload controls in such a situation is implementation-dependent, with an option of simply discarding messages in excess of the signaling transfer point (STP) processing capability. It is this option that is studied. Call completion performance, rather than message throughput is considered as being the primary performance measure of interest since it most accurately reflects the service provided to customers. To determine realistic call completion estimates. the authors explicitly incorporate into their analysis the effects of application-level recovery procedures and customer reattempts, both of which significantly impact the service levels achieved. In so doing, they demonstrate that message throughput can be a very misleading measure of the network's ability to provide service. The need for some form of feedback mechanism to the traffic sources that will allow them to appropriately control traffic entering the network is demonstrated     

Response-time evaluation of transaction-oriented applications inVSAT networks

The authors present modeling of an integrated terrestrial-satellite network using very small aperture terminals (VSATs), for transaction-oriented computer applications. They address the analytical and simulation modeling techniques used to study the network performance. More specifically, these models are used to obtain network response times for an integrated terrestrial-VSAT network for inquiry-response applications under various load conditions at the cluster controller units. The major result obtained through this analytical and simulation modeling suggests that communications satellite channels can be used efficiently for the interactive inquiry-response traffic at an acceptable response time, provided that the system parameters including the link transmission rates are chosen appropriately     

基于风险估计模型的网格任务动态级调度算法

网格中的各种资源的失效是不可避免的,为尽可能减少由于资源失效导致任务执行失败带来的影响,网格的任务调度算法的目标不仅要最小化任务执行的时间,还要兼顾考虑任务在资源上执行失败的风险。提出了Risk—DLS（Dynamic Level Scheduling）算法是将资源的风险估计模型与DLS算法相结合,通过仿真试验与DLS算法相比较,新算法在选择适当的参数的情况下,不仅能够最小化DAG型应用的完成时间,还能提高任务执行的成功率,有效的减少网格环境的不确定性对任务执行的影响。     

End-to-end service level agreements for complex ICT solutions

The management of converged solutions for corporate customers presents service providers with substantial challenges in the delivery of service level agreements (SLAs). While historically SLAs have focused on individual network products, corporate-scale customers are increasingly demanding SLAs targeted at their business applications. The goal is quite simply to guarantee that the applications will work — with predictable and consistent performance over the entire IT and networking infrastructure. Underlying this, however, is the complex myriad of information and communications technology components from the desktop through to the data centre, each of which contribute to the overall performance. This paper summarises the issues and requirements for such application-level SLAs, and outlines a five-step technology roadmap to achieve a sophisticated level of capability.     

An 802.11 MAC layer covert channel

For extremely sensitive applications, it may be advantageous for users to transmit certain types of data covertly over the network. This provides an additional layer of security to that provided by the different layers of the protocol stack. In this paper we present a covert side channel that uses the 802.11 MAC rate switching protocol. The covert channel provides a general method to hide communications within currently deployed 802.11 LANs. The technique uses a one‐time password (OTP) algorithm to ensure high‐entropy randomness of the covert messages. We investigate how the covert side channel affects network throughput under various rate‐switching conditions with UDP‐based and TCP‐based application traffic. We also investigate the covertness of the covert side channel using standardized entropy. The theoretical analysis shows that the maximum covert channel bandwidth is 60 bps. The simulation results show that the impact on network throughput is minimal and increases slightly as the covert channel bandwidth increases. We further show that the channel has 100% accuracy with minimal impact on rate switching entropy for scenarios where rate switching normally occurs. Finally, we present two applications for the covert channel: covert authentication and covert WiFi botnets. Copyright © 2010 John Wiley & Sons, Ltd.     

基于深度报文检测和深度流检测的骨干网流量特征分析

Based on the massive data collected with a passive network monitoring equipment placed in China's backbone, we present a deep insight into the network backbone traffic and evaluate various ways for improving traffic classifying efficiency in this paper. In particular, the study has scrutinized the network traffic in terms of protocol types and signatures, flow length, and port distribution, from which meaningful and interesting insights on the current Internet of China from the perspective of both the packet and flow levels are derived. We show that the classification efficiency can be greatly improved by using the information of preferred ports of the network applications. Quantitatively, we find two traffic duration thresholds, with which 40% of TCP flows and 70% of UDP flows can be excluded from classification processing while the impact on classification accuracy is trivial, i.e., the classification accuracy can still reach a high level by saving 85% of the resources.     

Topology discovery services for monitoring the global grid

The dynamic joint optimization of both computational and network resources has the potential of guaranteeing optimal performance Io geographically distributed grid applications. A grid network information and monitoring service (NIMS) has been recently proposed to complement computational resource status information with network resource status information. NIMS information includes, but is not limited to, information already available in the network control plane (e.g., network topology, link capacity occupation, communication delay). This study first reviews some measurement methodologies and network sensors suitable for implementing NIMS components, and then describes some tools currently utilized for monitoring grid network infrastructures. Finally, two implementations of a NIMS component, called the topology discovery service (TDS), are proposed and evaluated. The TDS provides grid users (e.g., applications) or the programming environment middleware with up-to-date information on the grid network infrastructure topology and status. Both proposed implementations can be utilized in any global grid network based on commercial routers without requiring modifications of router management and control protocols.     

Time-sensitive software-defined networking (Ts-SDN) control architecture for flexi-grid optical networks with data center application

Data center interconnected by flexi-grid optical networks is a promising scenario to meet the high burstiness and high bandwidth requirement of data center application, because flexi-grid optical networks can allocate spectral resources for applications in a dynamic, tunable and efficient control manner. Meanwhile, as centralized control architecture, the software-defined networking (SDN) enabled by OpenFlow protocol can provide maximum flexibility for the networks and make a unified control over various resources for the joint optimization of data center and network resources. Time factor is firstly introduced into SDN-based control architecture for flexi-grid optical networks supporting data center application. A traffic model considering time factor is proposed, and a requirement parameter, i.e., bandwidth-period product is adopted for the service requirement measurement. Then, time-sensitive software-defined networking (Ts-SDN)-based control architecture is designed with OpenFlow protocol extension. A novel deadline-driven PCE algorithm is proposed for the deadline-driven service under Ts-SDN-based control architecture, which can complete data center selection, path computation and bandwidth resource allocation. Finally, simulation results shows that our proposed Ts-SDN control architecture and deadline-driven PCE algorithm can improve the application and network performance to a large extent in blocking probability.     

A Resource Allocation Approach for the Generation of Service-Dependent Demand Matrices for Communications Networks

Estimating point-to-point demands from partially available information, such as total demand volumes originating and terminating at nodes and traffic volumes routed on links, has significant applications in various areas, such as communications network planning and transportation planning. Existing methods include matrix and link scaling methods, statistical methods, more complex mathematical programming models, and forecasting using demographic data. We present a new mathematical programming model based on equitable resource allocation. The model considers multiple services, e.g., data, video, and voice, and generates a point-to-point demand matrix for each service. Originating and terminating demands for each service and link loads, aggregated over all services, are viewed as resources. Each point-to-point demand is associated with a performance function that measures its weighted, normalized deviation from a target defined by a service-dependent community of interest matrix. The model formulation has a lexicographic minimax objective function and multiple knapsack resource constraints. The model has an intuitively appealing interpretation and a specialized algorithm can generate demand matrices for large network problems very fast.     

面向大规模网络的聚集TCP流量模拟方法研究

为平衡大规模网络骨干网上TCP流量模拟的准确性和复杂度,提出模拟接入网出口处的TCP聚集流量,而不从单台主机或会话的层面进行流量模拟。从网络流量的自相似性出发,提出基于应用层和网络传输层自相似的TCP聚集流量模型:在应用层通过聚集多个服从Pareto分布的ON/OFF源模拟生成自相似流量,在网络传输层根据TCP机制实现对应用层流量的聚集传输控制。模拟结果显示,在流量负载比例、流量自相似性以及骨干节点丢包率等方面,模拟结果与采集自某互联网交换中心的实际数据吻合良好,表明了该模型的有效性。     

Study of the relationship between peer-to-peer systems and IP multicasting

Peer-to-peer (P2P) systems, which have attracted public attention, involve a number of directly connected "peers" exchanging various types of information among themselves. However, applications based on P2P systems generally generate a lot of traffic and require not only the resources of every peer (e.g., CPU, memory, and bandwidth) but also many network resources. We think the main problems of P2P systems can be solved using IP multicasting, which is becoming popular as an effective way to transfer streaming media over the network and can reduce network traffic and the load on a streaming server. We propose applying IP multicasting to P2P systems. After mentioning that almost every application can be improved by our proposal, we discuss which routing protocol is suitable. Finally, we show that we can get good results even when part of the network does not support IP multicasting.     

Signaling architectures and recovery time scaling for grid applications in IST Project MUPBED

It is now evident to the research community that local computational resources cannot keep up in an economical way with the demands generated by some users/applications. Therefore, distributed computing and the concept of a computational grid are now emerging. Novel transport network concepts are needed to support such visions, and high-speed intelligent optical networking may be the required infrastructure that will enable global grids. Emerging utility grid applications like business continuity and disaster recovery have strong requirements on the dynamic optical networks connecting the distributed grid resources. Supporting grid networking with an intelligent optical network (ION) infrastructure will allow utility grid applications the necessary flexibility with the required QoS (e.g., high bandwidth, reliability, limited delay). Emerging QoS requirements, such as scalable recovery time, highly depend on the ION's signaling architecture. This article gives simple analytical models for the implementation options of optical control plane signaling, shows simulation models for different resilience strategies, and offers some illustrative numerical comparisons to support the aforementioned efforts. This research area is also discussed, among others, in the European research project Multi-Partner European Testbeds for Research Networking (MUPBED).     

Performance Modeling of an Integrated Wireless Network Using WiMAX as Backhaul Support for WiFi Traffic

We propose an analytic model for an integrated wireless network using WiMAX as backhaul support for WiFi traffic and evaluate the system performance. A unique feature for the proposed model is that the WiFi traffic completely reflects the realistic user mobility. A WiFi call can be overflowed to its overlaid WiMAX cell when it is rejected at the WiFi cell; a WiFi call may also work for some period of time in the WiFi cell and then make a vertical handoff to its overlaid WiMAX cell when it wants to move from its current WiFi cell (e.g., office) to its target WiFi cell (e.g., airport). Further, the target WiFi cell may be located at another place in the same WiMAX cell, or at a different WiMAX cell. We use Markov processes to model the dynamics of the WiMAX traffic and WiFi traffic including the overflowed WiFi traffic, the vertical handoff WiFi traffic, the horizontal handoff WiFi traffic, and the take-back WiFi traffic. We derive the explicit expressions of various traffic arrival rates and performance metrics and analyze the performance improvement of the WiFi traffic and the impact on the WiMAX traffic due to backhaul support. Numerical results are provided for further understanding of the gain and loss of the integrated architecture.     

Dynamic priority re-allocation scheme for quality of service in IEEE 802.11e wireless networks

In IEEE 802.11e Enhanced Distributed Coordinator Function (EDCF) (In the recently approved IEEE 802.11e standard, EDCF is renamed to enhanced distributed channel access (EDCA). Throughout this paper, we use EDCF for consistency with early work in the literature.), per-flow service differentiation is achieved by maintaining separate queues for different traffic categories (TCs). However, due to its static Quality of Service (QoS) parameter setting, EDCF does not perform adequately under high traffic load (Romdhani et al., Proceedings of IEEE wireless communications and networking conference, 2003). We present an extended performance model of EDCF and analyze conditions for network getting overloaded. With this extended model, we show that the overall throughput of a network can be improved by changing the distribution of the number of active stations (an active station is one that has a pending packet to be sent) over a set of TCs. Hence, we propose to dynamically re-allocate flow priorities evenly in order to maintain high system performance while providing QoS guarantee for individual real-time flows. Our scheme has several interesting features: (1) performance of EDCF is improved; (2) low priority flows are not starved under high traffic load; (3) misuse of priority (misuse of priority means that a flow requests much higher priority than necessary) can be easily handled. Simulations are conducted for both infrastructure-based and Ad hoc models. Results show that dynamic priority re-allocation does not decrease throughput of real-time flows under low to medium loads, while considerable improvement over EDCF is obtained even under high loads, making it easy to support multimedia applications.     

Network issues for wireless communications

This article specifically focuses on wireless personal communications, i.e., wireless access, that provides either terminal or personal mobility. In particular, we discuss some important issues in networking, traffic, and performance. Although within radio and networking aspects there are significant commonalities between traditional cellular mobile communications and wireless personal communications, there exist distinct differences due to radio propagation and fading effects, interference environment, smaller cell sizes, type and pattern of mobility, and call delivery. Indeed, with respect to networking issues, a large set of system choices, characteristics of traffic to be carried, and important parameters have to be considered. These include the problems involved in selecting an appropriate multiaccess technology to efficiently handle the required subscriber service profile across a multiplicity of systems to complete a call. To present a meaningful discussion of these issues, we address in some detail radio resource assignment, mobility management, call control, and traffic aspect, which have significant impact on the network performance