一种基于QoS的云负载均衡机制的研究

提出一种基于QoS的云负载均衡机制,即:构建QoS模型和云资源模型;建立资源度量与QoS属性之间的映射;对虚拟机实例负载状况和虚拟机集群资源利用状况进行量化评估;感知用户的QoS并对比所监控的云节点的资源度量情况,根据对比结果,通过任务调度算法和弹性伸缩算法分别实现任务的分发和虚拟机集群的弹性伸缩,最终达到优化的负载均衡的目的.通过模拟试验,结果表明本方法与Round robin算法相比,有更好的负载均衡效果.     

分布式视频服务器的动态数据分配算法

针对分布式视频点播系统，提出了三种改进的动态数据分配算法：DRR(Dynamic Round Robin video data allocation)，DBWP(Dynamic Bandwidth Weighted Partition video data allocation)，DPB(Dynamic Popularity Based video data allocation)。仿真结果表明：分别与RR，BWP，PB算法相比，DRR，DBWP和DPB算法具有较好的环境适应性，特别是在视频对象访问频率及视频服务器带宽差异较大、环境扰动增强的情况下，DBWP和DPB算法可明显降低分布式视频服务器的视频对象访问响应时间。     

SDN控制器部署和基于流的动态管理方案*

软件定义网络（Software Defined Networking, SDN）通过构建独立的控制平面,极大地降低了网络设备及管理的复杂性。但在大规模广域网部署中,这种逻辑集中的方法在性能和扩展性方面存在诸多限制。因此控制平面多控制器的部署是一个非常重要的任务,其可以通过配置有限的资源来满足多样化需求。这些需求包括延迟限制、容错能力和负载均衡。本文首先提出一种控制器部署方法,该方法用于在给定的网络拓扑中完成控制器位置的部署。其次针对该部署方法设计了两种算法：交换机迁移算法将过载控制器域内交换机迁移到未过载控制器管理域中去,实现控制器间负载均衡;控制器池伸缩容量算法（Controller pool Scalable Capacity Algorithm, CSCA）实现池内控制器数目的动态伸缩,提高网络资源利用率。最后,仿真结果表明,相较LG方法,该方案在完成部署任务的同时,能够减少系统管理开销。     

Massive MIMO系统基于子空间的半盲信道估计

针对大规模MIMO系统中存在的导频污染问题,结合目前研究的基于奇异值（SVD）分解的信道估计算法,在考虑到该算法中的协方差矩阵是用有限的样本数据代替真实数据必然存在偏差的问题,给出了一种联合ILSP（Iterative Least Square with Projection）的基于SVD的半盲信道估计算法。仿真结果表明改进后的信道估计算法能够有效减小已有算法中存在的偏差问题,提高信道估计精确度,有效减轻导频污染给大规模MIMO系统带来的影响,从而实现大规模MIMO系统性能的提升。     

基于模拟退火算法的资源负载均衡方案

针对大型ERP系统的网络资源负载均衡问题展开研究。在理论上对问题的需求进行建模分析,设计了满足主机和网络性能约束的启发式目标函数,将模型转化为度约束最小生成树问题,设计了一种模拟退火算法对此问题进行处理,提出了基于该算法的资源负载均衡方案LABS。理论分析表明,整个网络执行该方案的时间复杂度为节点规模的平方阶,说明方案具有较强的可用性与可伸缩性。仿真实验结果显示,通过选择适当的启发因子,算法不仅可以吸纳大部分节点协同参与负载均衡操作,还能够显著减少系统中的瓶颈节点数,降低平均资源使用率。     

基于NS-2的自适应差额IEEE 802.11 PCF轮询机制

在研究无线局域网IEEE 802.11标准点协调功能PCF(Point Coordination Function)机制不足的情况下,针对其中Round Robin轮询机制存在的QoS(Quality of Service)不能得到有效保障的问题,提出了一套自适应差额轮询调度算法解决方案,阐述了算法的核心思想,介绍了调度机制的执行过程,并且通过修改PCF源代码,在网络模拟器NS-2平台上对算法进行了脚本模拟.实验结果表明,自适应差额IEEE 802.11 PCF轮询算法比传统的Round Robin算法在端到端延迟、系统吞吐量和包延迟等QoS服务性能方面都有显著的提高.     

改进的变步长频域批处理LMS算法

现有的单载波频域均衡技术中的定步长频域批处理LMS（Frequency-Domain Block Least Mean Square,FBLMS）算法,在收敛速度和稳态误差之间存在矛盾。针对这个问题,基于对变步长LMS算法的研究分析,提出了一种新的改进的变步长频域批处理LMS自适应滤波算法,通过变步长因子以及频域权系数抽头泄漏能很好地协调收敛速度和稳态误差之间的矛盾,并且还具有较低的算法复杂度的特点。通过Matlab对提出的新算法进行计算机仿真验证,结果表明该算法有较好的收敛速度和较小的稳态误差。     

基于区域划分的局部更新指纹定位算法

针对室内定位指纹数据库更新成本过高的问题,设计了一种通过区域划分进行局部更新指纹数据库的RFID（Radio Frequency Identification,射频识别技术）室内定位算法。该算法通过聚类算法将指纹地图分成若干个子区域,每个子区域选取一个代表点代表该子区域的指纹有效性,通过检测代表点的有效性来选择加权k近邻算法（Weighted k-Nearest Neighbor,WkNN）定位或子区域数据库的局部更新。实验结果表明,该算法在低成本的条件下极大限度地提高了定位精度和长期定位稳定性。     

一种有效的Web负载均衡器的设计与实现

随着Web应用技术大量地应用到各中小企业的信息系统中，迫切需要一个能够提供良好性能价格比的Web服务器负载均衡系统。一方面由于Round Robin DNS技术的局限性达不到企业的要求，另一方面由于硬件负载均衡器的昂贵费用使中小企业无法承受。该研究采用J2EE平台的Servlet技术提供的Filter功能，设计并实现了一个比较适合中小企业的Web服务器负载均衡系统框架。     

异构多核处理器的任务调度算法

在研究Min-min、Max-min算法和Sufferage算法基础上,针对异构多核处理器的特点,提出一种任务静态调度算法——自适应分段Sufferage算法（Adaptive Segmented Sufferage,ASS）。该算法以最早完成时间和负载均衡为目标进行任务分配,先将任务分配分成两个阶段：在第一个阶段以最少完成时间作为分配原则进行分配,选择单位时间内节省时间最多的任务先分配;在第二个阶段以负载均衡为分配原则进行分配,选择执行时间大的任务先分配。然后选取不同调节参数,对任务进行多次重新分配,以最小的最大完成时间为最后分配结果,实现自适应调节。通过实验验证,该算法在实现最少完成时间的前提下能很好地达到负载均衡。     

加权轮询算法在TCP非对称网络中的应用及其性能分析

ＴＣＰ是利用接收端返回的确认信息调整发送端数据发送的传输层协议。在非对称信道存在的网络中,当传输确认信息的信道带宽远小于传输数据的信道带宽,确认信息的传输速率小于数据包的传输速率时,ＴＣＰ数据包的传输受到确认信息返回速率的限制。如何在数据包和确认包共享同一个上行信道的情况下,能够既保证上行信道上数据包传输带宽,又不使下行信道上的数据包传输受到抑制是笔者所关心的问题。文章在上行信道采用了一种加权轮询分组调度算法对其数据包和确认包的发送进行控制来解决这一问题,并通过理论分析对该分组调度算法的性能进行分析,最后用仿真验证了对该算法的分析。     

An FEC scheme combined with weighted scheduling to reduce multicast packet loss in IPTV over PON

Congestion is one of the most important challenges in optical networks. In a Passive Optical Network (PON), the Optical Line Terminal (OLT) is a bottleneck and congestion prone. In this paper, a framework is proposed with Forward Error Correction (FEC) at the IP layer combined with Weighted Round Robin (WRR) at the scheduling level to overcome packet-loss due to congestion in the OLT in order to achieve efficient video multicasting over PON. In the FEC scheme, Reed-Solomon (RS(n,k)) with erasure coding is used, where (n−k) erroneous symbols per n symbol blocks can be corrected. In our framework, an Internet Protocol TeleVision (IPTV) service provider uses the mentioned RS coding and generates redundant packets from regular IPTV packets in such a way that an Optical Network Unit (ONU) can recover lost packets from received packets, thus resulting in a better video quality. Simulation results show that using the proposed framework, an ONU can recover many lost packets and achieve better video quality under different traffic loads for its users. For instance, the proposed method can reduce packet loss rate by almost 55% and 10% under traffic load 0.9, respectively, compared with the Round Robin (RR) and WRR methods under symmetric traffic load. When High Receivers Queue (HRQ) traffic (i.e., traffic received by many users) is twice Low Receivers Queue (LRQ) traffic (i.e., traffic received by a small number of users), this reduction is almost 86% and 30% under traffic load 0.9. Finally, when LRQ traffic is twice HRQ traffic, the reduction in packet loss rate is almost 70% and 91% at traffic load 0.5.     

Analyzing Weighted Round Robin policies with a stochastic comparison approach

We study queuing delays for Weighted Round Robin (WRR) scheduling policies. The delay characteristics of these policies can be estimated through worst-case bounds as other fair queuing scheduling policies. However, these deterministic bounds are in general not accurate and they require that the input process has been shaped. Under Markovian arrival hypothesis, the delay of WRR   policies can be evaluated through Markovian numerical analysis. Nevertheless, this analysis is limited to small parameter sizes because of the state space explosion problem. We propose to apply the stochastic comparison approach. We build a bounding model based on the aggregation of sessions, and we show that under the same arrivals, the packet delays in the bounding model are larger in the _?st${?}_{\mathrm{st}}$ stochastic order sense than the packet delays in the original model. Due to this aggregation, the state space size is drastically reduced and we are able to provide stochastic delay bounds for WRR policies. We discuss the accuracy of the proposed bounds through numerical examples.     

无线ad hoc网络中区分服务存在的问题研究

对无线ad hoc网络中的区分服务进行了仿真,发现区分服务机制为业务流分配的带宽资源与其优先级调度机制的权值比例并不总是保持一致。结合仿真记录文件进行分析,得出造成这一现象的原因是由于受到MAC机制具有带宽分配的能力。最后指出,在无线ad hoc网络中,单纯依靠网络层的区分服务机制并不能实现成比例的分配带宽。     

基于FPGA的轮询合路的设计和实现

针对高密度接口设计中基于字节处理和整包处理的转换问题,本文提出了分片轮询调度和改进式欠账轮询调度相结合的调度策略,该策略在很大程度上保证了公平性和稳定性。仿真结果显示,该设计完全符合要求。     

基于轮循机制和RED的语音流调度机制

本文提出了一种新的VoIP业务流调度机制(RR-RED),通过随机早期检测(RED)和轮循机制(Round Robin)控制主动丢包。该机制很好的继承了RED算法的优点,同时根据VoIP业务流的特点,降低了连续丢包率,同时通过提高多业务流丢包的公平性,提高了链路利用率。     

Relative Time Quantum-based Enhancements in Round Robin Scheduling

Modern human life is heavily dependent on computing systems and one of the core components affecting the performance of these systems is underlying operating system. Operating systems need to be upgraded to match the needs of modern-day systems relying on Internet of Things, Fog computing and Mobile based applications. The scheduling algorithm of the operating system dictates that how the resources will be allocated to the processes and the Round Robin algorithm (RR) has been widely used for it. The intent of this study is to ameliorate RR scheduling algorithm to optimize task scheduling. We have carried out an experimental study where we have developed four variations of RR, each algorithm considers three-time quanta and the performance of these variations was compared with the RR algorithm, and results highlighted that these variations performed better than conventional RR algorithm. In the future, we intend to develop an automated scheduler that can determine optimal algorithm based on the current set of processes and will allocate time quantum to the processes intelligently at the run time. This way the task performance of modern-day systems can be improved to make them more efficient.     

Approximate sorting of packet-scheduling in high-speed networks

Fairness,latency and computational complexity are three important factors in evaluating the performance of a scheduling algorithm.Fairness must be satisfied so that service can be distributed according to the reserved rate.Only when latency is irrelevant to the number of connections,is it possible to minimize the end-to-end delay through controlling the reserved rate.Among existing scheduling algorithms,Round Robin is the least complex.However,conventional Round Robin is unable to ensure fairness,and the improved round robin algorithms like Deficit Round Robin,Weighted Round Robin and VIrtual Round Robin are unable to ensure that their latencies are irrelevant to the number of connections although they guarantee fairness,Potential Round Robin developed for analysis of fairness and latency reduction is thus propsed.It is based on the introduction of a new concept,Round Potential Function.The function splits service time into a number of service round periods to guarantee fairness regardless of the serving process used in the period.In the analysis of latency,service round periods are re-split into multiple scanning cycles for further service distribution with approximate sorting between scanning cycles,As a result,latency is no longer relevant to the number of connections while the low complexity of round robin is kept.     

加权最小二乘估计在无线传感器网络定位中的应用*

节点自身定位是目前无线传感器网络领域研究的重点之一,定位误差累积问题是节点定位中必须解决的一个关键问题,利用加权最小二乘估计的方法可以有效抑制累积误差的影响。介绍了如何将加权最小二乘估计应用于节点定位以及如何合理地选择加权系数以降低定位误差。仿真实验表明,运用加权最小二乘估计可以有效地抑制误差累积的影响,提高定位精度。