基于两级LRU机制的大流检测算法

高速网络中及时准确地识别大流量对象对网络测量、控制和管理有重要的意义.该文提出了一种基于两级LRU机制的大流检测算法并分析了新算法中的参数与平均误差之间的关系.新算法通过增加一级用于过滤小流的LRU提高测量准确性.算法分析表明:新算法具有10Gbps的线速处理能力.该文基于实际互联网数据进行了实验对比,结果显示:与现有算法相比,新算法具有更高的测量准确性和实用性.     

一种基于大小流区分计数的公平抽样算法

针对一种草图指导公平抽样(SGS)算法对小流估计误差大的问题,该文提出一种基于大小流区分计数的包公平抽样算法(DCMFS),并给出哈希冲突对SGS算法估计误差影响的定量分析结果。DCMFS采用大小流区分计数器,对小流采用逐流精确计数,对大流采用哈希计数。理论分析及实际的数据仿真结果均表明,DCMFS算法对小流能够实现逐流精确统计,对大流的估计标准差接近公平抽样估计标准差理论值上限。算法采用不等长位宽计数器结构,保证其空间复杂度较SGS和自适应非线性抽样方法(ANLS)没有增加;引入计数器置换使得算法时间复杂度略有提高,但仍能满足10 Gbps线速处理要求。     

基于LRU的大流检测算法

高速网络中,检测大流是进行准确流量测量的一种重要可扩展解决方案。该文提出了一种新的基于LRU大流检测算法。它通过引入小流早期丢弃和大流预保护机制以提高测量准确性。算法分析表明:新算法具有10Gbps线速处理能力。该文基于实际互联网数据进行了实验比较,结果显示:与已有算法相比,新算法具有更高的测量准确性和实用性。     

基于抽样流长与完全抽样阈值的异常流自适应抽样算法

高速IP网络的流量测量与异常检测是网络测量领域研究的热点。针对目前网络流量测量算法对小流估计精度偏低,对异常流量筛选能力较差的缺陷,该文提出一种基于业务流已抽样长度与完全抽样阈值S的自适应流抽样算法(AFPT)。AFPT算法根据完全抽样阈值S筛选对异常流量敏感相关的小流,同时根据业务流已抽样长度自适应调整抽样概率。仿真和实验结果表明,AFPT算法的估计误差与理论上界相符,具有较强的异常流量筛选能力,能够有效提高异常检测算法的准确率。     

基于多位域的等概率随机IP流抽样算法

对IP流信息的全方位提取有助于实现网络实时监控,精细管理,有利于网络安全性能的提升。已有的等概率随机IP流抽样算法将大量的IP流重复抽样,浪费了宝贵的计算和存储资源。针对这个问题,在原有算法的基础上设计了一种新的等概率随机IP流抽样算法,该算法在Bloom Filter的基础上采用三层位域,两层同时测量,结果取交集的方法,便于实际使用并且有效减少了已被抽样的IP流被重复抽样。实验结果表明:新方法能够大幅度提高测量精度,节约了系统资源,可以适用于10 Gb/s左右的高速网络之中。     

基于Octeon多核网络处理器的IPv6联动IPS研究与设计

对基于Octeon多核网络处理器的新一代IPv6高速网络联动入侵防御系统进行研究,设计了新型联动入侵防御原型.系统基于Octeon多核的高速处理,并结合了IPv6网络中入侵的新特点.在基于入侵检测规则库规则匹配技术的基础上,运用新型的协议分析技术和基于流的检测技术,在Octeon多核间分配控制层与数据层的不同执行,采用命名块机制进行多核间通信,通过数据层核向控制层核的反馈,实现了流处理及协议分析模块与控制模块的高速联动.系统实现了Gbps级的高速入侵检测与联动防御处理.     

Netronome：为亚太区量身定制解决方案

Netronome在日前举办的英特尔信息技术峰会（IDF）上,展出和演示了网络流引擎（NFE）加速卡和网络流处理平台解决方案,并指出针对中国和亚洲网络市场的特定需求和应用所提供全面解决方案。互联网的快速发展带来了运营商和企业网络流量的持续高速上升,推动今天10Gbps的带宽需求迅速增长到40Gbps以上。     

高速电路信号完整性的测试方法及分析

在高速数字系统设计中,信号完整性问题是必须慎重考虑的问题.该文就1Gbps以上的高速信号,对其信号完整性的测试方法进行了详细论述,包括TDR阻抗测试、眼图测试和串扰测试等.运用这些方法,给出了一个3.125Gbps背板的实际测试结果,并针对两种不同的测试设备进行了对比分析.     

基于抽样和哈希技术的长流测量算法研究

在互联网中理解网络行为最高效的途径即是对网络数据流量进行检测与分析,它是对已有互联网的组建、规范化和改造的依据,同时也是对Internet进行检测的重要环节.为了解决网络中的资源和高速IP流量之间的冲突问题,需要对网络流进行多种方式的处理与算法研究.本文首先提出了改进的数据抽样技术并综合论述了现阶段基于抽样技术的数据测量算法的研究.同时通过对重要数据参数的重新设置和分析,并结合使用多种数据取样的方法,探讨改进的数据空间映射技术与现阶段的各种取样方式在测量网络长流算法中的综合应用.     

基于洗牌策略的Sybil攻击防御

 洗牌策略从理论上解决了分布式哈希表(DHT)的Sybil攻击问题.为克服敌手作弊,引入受信节点构成分布式认证系统,由受信节点对新加入节点进行认证,保证节点签名和ID不能伪造;同时引入记录洗牌加入过程的票据来判定节点合法性,杜绝了敌手积累过期ID.由于保存票据的数量决定了论文提出算法的应用效果,通过理论分析和仿真实验证实设计的算法需要保存的票据数量不大,保证了算法的可行性.     

基于压缩感知和最小二乘的分布式协作频谱感知

针对认知无线电(CR)集中式频谱感知算法对融合中心要求高,而且对节点失效的容忍性也不高等缺点,提出了一种基于压缩感知的分布式多节点协作算法.认知无线电网络中每个CR节点在接收信号频谱后,首先根据压缩采样理论在本地获取压缩采样测量值,然后利用l1范数约束的最小二乘算法在本节点估计频谱,把在此节点估计的频谱传给下一相邻节点,以此进行迭代优化直到算法收敛.理论分析和仿真结果表明,所提算法不仅计算复杂度低,收敛速度快,而且精确重构性能好,可靠性较高.     

Algorithm of conditional minimization of the goal function for optimal routing in information networks

Anew algorithm has been proposed for solving the optimal routing problem. This algorithm is based on applying the Kirchhoff laws to information networks and does not require the mandatory use of derivatives of the goal function making it quite convenient for distributed realizations. The algorithm convergence is substantiated by drawing an analogy between information and electric networks. On the basis of a case study of the network it was shown that its speed is tens of times as high as that of the flow deviation algorithm. It was shown that theoretical labor intensity of implementing this method is substantially less than that of the algorithms based on finding the shortest routes, since the cyclic part of this algorithm does not contain laborious logical operations.     

Quality‐of‐Service Mechanisms for Flow‐Based Routers

In this paper, we propose quality of service mechanisms for flow‐based routers which have to handle several million flows at wire speed in high‐speed networks. Traffic management mechanisms are proposed for guaranteed traffic and non‐guaranteed traffic separately, and then the effective harmonization of the two mechanisms is introduced for real networks in which both traffic types are mixed together. A simple non‐work‐conserving fair queuing algorithm is proposed for guaranteed traffic, and an adaptive flow‐based random early drop algorithm is proposed for non‐guaranteed traffic. Based on that basic architecture, we propose a dynamic traffic identification method to dynamically prioritize traffic according to the traffic characteristics of applications. In a high‐speed router system, the dynamic traffic identification method could be a good alternative to deep packet inspection, which requires handling of the IP packet header and payload. Through numerical analysis, simulation, and a real system experiment, we demonstrate the performance of the proposed mechanisms.     

基于射线跟踪的车联网毫米波信道准平稳区间研究

针对28 GHz车联网中车对基础设施（Vehicle-to-Infrastructure，V2I）毫米波信道，开展了准平稳区间的研究.首先，分析了信道准平稳区间的理论基础和计算方法.其次，利用功率相关（Correlation of Power，CP）算法和谱散度（Spectral Divergence，SD）算法进行准平稳区间的计算.结果表明，CP算法复杂度低且精度较高，更适合用于计算V2I通信场景下的准平稳区间.最后，利用CP算法计算出的准平稳区间作为新的采样间隔，从而减少了射线跟踪仿真器的采样点数.通过原始采样和减少采样点后的信道特性的比较，在合理的门限值下，利用信道准平稳区间的方法加速了射线跟踪仿真器，并且不会造成信道参数的失真.另外，对于信道特性的分析结果也对车联网在毫米波频段的仿真和设计具有重要意义.     

基于带通采样QPSK高速解调器的技术分析

首先通过对带通采样定理的分析,提出了正交相移键控(QPSK)高速解调器的实现方案,着重对带通采样解调算法中时钟的选取、并行匹配滤波算法、载波恢复算法和定时估计算法进行了详细描述,并用硬件编程语言编写了整个算法的代码,最后在硬件平台上下载了代码文件,验证了解调性能,测试结果表明,采用该算法实现的高速解调器在性能上有了很大的提高。     

Online end‐to‐end quality of service monitoring for service level agreement management

A major challenge in network and service level agreement (SLA) management is to provide Quality of Service (QoS) demanded by heterogeneous network applications. Online QoS monitoring plays an important role in the process by providing objective measurements that can be used for improving network design, troubleshooting and management. Online QoS monitoring becomes increasingly difficult and complex due to the rapid expansion of the Internet and the dramatic increase in the speed of network. Sampling techniques have been explored as a means to reduce the difficulty and complexity of measurement. In this paper, we investigate several major sampling techniques, i.e. systematic sampling, simple random sampling and stratified sampling. Performance analysis is conducted on these techniques. It is shown that stratified sampling with optimum allocation has the best performance. However, stratified sampling with optimum allocation requires additional statistics usually not available for real‐time applications. An adaptive stratified sampling algorithm is proposed to solve the problem. Both theoretical analysis and simulation show that the proposed adaptive stratified sampling algorithm outperforms other sampling techniques and achieves a performance comparable to stratified sampling with optimum allocation. A QoS monitoring software using the aforementioned sampling techniques is designed and tested in various real networks. Copyright © 2007 John Wiley & Sons, Ltd.     

基于实时数字信号处理的宽带单比特瞬时测频接收技术

针对现代电子战接收机的大瞬时带宽、实时处理、高灵敏度、快速测频等高性能指标要求,提出了一种基于实时数字信号处理的宽带单比特瞬时测频接收机方案,结合大带宽高速采样的实时快速处理,将单比特接收算法、滑动快速傅里叶变换(FFT)算法、瞬时频率测量算法应用于该算法结构,给出了超高速单比特采样和高速实时信号处理的硬件实现。仿真结果表明,该接收机不仅具备大带宽、实时处理、高频率分辨率等特性,同时具有优异的信号检测灵敏度。     

Multi-Hop WiMAX Networks: A New Opportunistic Algorithm Using AMC and PUSC Mode for Different Mobility Users

Resource allocation algorithms play the main role in provisioning high throughput in broad band wireless communications. The previous studies in WiMAX networks have considered either AMC or PUSC techniques for the resource allocation algorithms. In this way, AMC based algorithms have presented high throughput for low speed users. Conversely, PUSC based algorithms have not been affected by users’ speed, whilst the throughput is low. To date, the importance of presenting proportional fair and maintaining the network throughput for the users with different speeds has not been acknowledged yet. This paper presents a novel opportunistic algorithm which is suitable for the users with different mobility speeds. The new algorithm uses two modes of AMC and PUSC simultaneously. The simulation results reveal that in comparison to similar algorithms, the proposed algorithm presents more proportional fair throughput when there are both high and low speed users.     

用于特定流匹配的随机矩阵映射Hash算法研究

针对常规的Hash算法用于流匹配时冲突率高且不可控制的缺点,提出了一种随机矩阵映射Hash算法。该算法通过预先优选一个随机数矩阵,然后将大集合的元素分块映射成随机矩阵中的元素,从而把一个大集合映射到一个小集合。测试结果表明,该算法运算速度快、空间利用率高、冲突率低,用于流匹配时匹配速度可以达到2Mpacket/s,支持规则数达5万条以上。     

组播路由调度的神经网络方法

本文探讨了在高速包交换计算机网络中，具有端到端时延及时延抖动限制的组播路由问题。首先给出了此类问题的网络模型及其数学描述，然后提出了基于Ｈｏｐｆｉｅｌｄ神经网络的组播路由优化算法。实验表明，本算法能根据组播应用对时延的要求，快速、有效地构造最优组播树，有较强的实时性。