基于输入队列的光分组交换短包抢先调度

光分组交换的输出队头阻塞引起分组的平均排队时延增加.分析了可变长分组的特点,提出了基于抢先方式的短包抢先调度(PSPP)算法,以减少分组在输入排队中的平均等待时间.在PSPP算法中,短包可以抢占长包的传输时间,获得优先的服务.分析和仿真结果表明,当到达业务负载为中或较低时,短包优先调度算法使短包的平均排队时延接近零,所有分组的平均等待时延减小,该算法还保证具有实时特性的TCP业务获得较低的平均等待时延.     

内容计费实时系统中一种提供比例时延区分的概率调度算法

针对OCS (online charging system)服务器,提出了一种新的请求调度算法,算法的基本思想是利用系统队列的长度、请求的到达率以及预先分配的时延区分参数作为调度的优先级依据,在调度时采用概率的方式选取需要服务的队列请求.实验结果表明,在考虑服务时延的情况下,新算法性能总是优于一些传统的PDD(proportional delay differentiation)调度算法,当请求服务时延增加和彼此差别很大时,性能优势相对更大,并且有效地满足了内容计费环境下提出的6点QoS(quality of service)要求.     

TWDM-PON中带有业务区分的混合资源调度算法

针对时分波分复用无源光网络长距离和广覆盖特性带来往返时延增大,造成高优先级业务时延性能严重恶化的问题,提出带有业务区分的混合资源调度算法.根据网络负载实时调整光网络单元组的波长数以实现资源共享,并动态划分子周期以对不同优先级业务进行区分服务;设计混合资源调度算法,以在线方式填充波长空隙,以离线方式保障资源高效调度.仿真结果表明,所提算法能够满足不同优先级业务的时延要求,保证服务质量,并提高信道利用率.     

智能网络存储系统负载均衡算法研究

负载均衡是大规模基于对象存储系统必须要考虑的重要问题.探讨了几种经典的负载均衡调度算法,介绍了它们的算法流程及其优缺点,最后根据INSS存储系统结构模型及其I/O请求流程,提出了一种二级动态反馈负载均衡模型,能最大程度的减少系统的平均总响应时间.     

一种HSDPA系统中依赖负载的正比公平调度算法

研究和仿真了HSDPA系统中分组调度算法(RR,Max C/I和PF)的系统吞吐量和用户公平性,仿真结果与理论分析表明在HSDPA系统中RR算法调度结果的公平性略逊于PF算法.为了补偿正比公平调度算法公平性,提出了依赖负载的正比公平(LDPF)调度算法.该算法旨在通过低负载时牺牲系统吞吐量换取用户间公平性.然后用扩展了爱立信EUPLANE模块的NS2网络仿真软件对LDPF算法进行了仿真.仿真结果表明LDPF调度算法在系统低负载时有效地补偿了先前的低速率用户,通过这种低负载和高负载时的时间分集,改善用户间的长时公平性.     

一种基于优先级队列的集群动态反馈调度算法

在分析现有面向LVS集群的负载均衡调度算法优缺点的基础上,提出了一种新的调度算法—基于优先级队列的动态反馈调度算法。该算法根据定期采集到的各服务器负载信息动态地调整各服务器的权值,并根据权值建立优先级调度队列借以实现连接的调度。算法可保证良好的负载均衡性,且时间复杂度降低至O(1)。     

一种内容中心网络中的热区控制及内容调度缓存算法

针对内容中心网络中同质化缓存造成的节点负载不均和存储资源无法有效利用的问题,提出一种热区控制及内容调度缓存算法.在内容请求时,根据节点介数与节点访问度综合判断节点热度,筛选出负载过重节点;缓存决策时,以流行度为依据将内容调度至空闲邻居节点,并设置生命期控制机制,从而达到分散请求、消除热区的目的.仿真结果表明该算法减少了请求时延与路由跳数,在提高缓存命中率的同时,有效改善了负载在节点上分布的均衡性.     

软件定义网络中面向时延和负载的多控制器放置策略

在多控制器管理的软件定义网络(SDN)中,时延和负载是控制器放置问题(CPP)要考虑的重要因素。该文以降低控制器之间的传播时延、流请求的传播时延和排队时延、均衡控制器间负载为目标,提出一种控制器放置及动态调整的策略,其中包括用于初始控制器放置的负载均衡算法(BCRA)和遗传算法(GA),用于动态调整控制器负载的在线调整算法(ADOA)。以上算法均考虑网络连通性。仿真结果表明:在初始控制器放置时,在保证流请求的传播时延、排队时延和控制器传播时延较低的情况下,BCRA部署在中小型网络中时,其负载均衡性能与GA相近且优于k-center和k-means算法;GA部署在大型网络中时,与BCRA, k-center和k-means算法相比,使得负载均衡率平均提高了49.7%。在动态情况下,与现有动态调整算法相比,ADOA可以保证较低排队时延和运行时间的同时,仍能使负载均衡参数小于1.54。     

5G网络切片场景中基于预测的虚拟网络功能动态部署算法

针对无线虚拟化网络在时间域上业务请求的动态变化和信息反馈时延导致虚拟资源分配的不合理,该文提出一种基于长短时记忆(LSTM)网络的流量感知算法,该算法通过服务功能链(SFC)的历史队列信息来预测未来负载状态。基于预测的结果,联合考虑虚拟网络功能(VNF)的调度问题和相应的计算资源分配问题,提出一种基于最大最小蚁群算法(MMACA)的虚拟网络功能动态部署方法,在满足未来队列不溢出的最低资源需求的前提下,采用按需分配的方式最大化计算资源利用率。仿真结果表明,该文提出的基于LSTM神经网络预测模型能够获得很好的预测效果,实现了网络的在线监测;基于MMACA的VNF部署方法有效降低了比特丢失率的同时也降低了整体VNF调度产生的平均端到端时延。     

FC-ABS调度算法及在NCS中的应用

针对网络控制系统(Network Control Systems, NCS)节点中的非周期任务,设计了一种FC-ABS调度算法.该调度算法可以根据非周期任务的时间特性采取不同的调度方式,并通过反馈调度减小非周期任务调度对周期任务的影响.仿真实验证明了算法的正确性.     

多核Web服务器中基于分配矩阵的动态请求调度算法

为了构建高性能的Web服务器,充分利用Web服务器中多核处理器的性能成为关键。传统的先到先服务策略没有考虑多核处理器的特点,不能充分利用多核处理器的性能。为解决此问题,该文提出一种基于分配矩阵的动态请求调度算法。该算法充分考虑了多核处理器的特点,可将同类动态请求动态分配至同一个处理器核心,提高了Web服务器处理动态请求的速度。仿真实验结果表明,采用该算法的Web服务器在自相似性、平均响应时间、丢包率等方面均优于传统的先到先服务算法。     

基于UWB的Ad hoc网络资源分配的研究

基于超宽带的Ad hoc网络的应用,提出了为多媒体业务提供QoS支持的自适应的资源分配方案。方案考虑了超宽带和Ad hoc网络的特性,同时也考虑了QoS的要求。提出的资源分配方案采取了业务区分和资源预留的方法,为链路分配网络资源（速率、功率）。根据吞吐量,丢包率和时延,对提出的方案做了详尽的性能评估,仿真结果证明了提出的方案的优越性。     

A preemption‐based scheduling algorithm for WiMAX networks

Several scheduling techniques were designed for the base station (BS) of IEEE 802.16e wireless interoperability of microwave access networks. However, depending on the BS scheduler alone to determine the servicing order of each connection might affect the accuracy of the scheduling process because the BS does not necessarily have enough up‐to‐date information about the current state of the connections at the subscriber station. In this paper, we propose a preemption‐based scheduling algorithm that focuses on improving the quality of service requirements of real‐time service flow classes. The proposed algorithm incorporates two schedulers, one at the BS and another one at the subscriber station. We have implemented and integrated the proposed algorithm with the network simulator NS2 using the Network and Distributed Systems Laboratory wireless interoperability of microwave access module. Simulation results have shown that the proposed approach outperforms other scheduling algorithms in terms of enhancing the throughput and the average delay of real‐time quality of service classes. Copyright © 2013 John Wiley & Sons, Ltd.     

On overload control of parlay application server in next generation network

After analyzing the features of the Parlay application server（AS）, which provides many services with different time constraints in the next generation network（NGN）, a selfadapted overload control algorithm implemented in Parlay AS is proposed. In this algorithm, the service process module in parlay AS uses the earliest deadline first （EDF） serving rule to serve messages, and the overload control mechanism, using the Leaky Bucket algorithm to control message flows, is triggered by estimating the queue waiting time of a newly arriving message. By simulation it shows that this algorithm is a simple, easy-to-use, and available algorithm.     

一种集群计算并行I/O中启发式文件分类分配策略研究

提出了一种新的应用于集群计算环境的文件分配算法,即启发式文件分类分配算法(Heuristic FileSorted Assignment algorithm,HFSA),它在保证系统负载基本均衡的前提下,按相似的访问服务时间对每个待分配的数据文件进行磁盘分配.具体实现时,首先对待分配文件按它们的服务时间大小进行排序,然后从集群中选择一个结点的磁盘,将有序的若干文件连续地分配到该磁盘上,直接该磁盘达到最大负载时为止.通过对新策略与已有的贪婪文件分配法进行实验比较,结果表明:在访问响应时间方面,系统轻负载时,缩短了20.2%,系统重负载时,缩短了31.6%,并且数据访问的速率越高,由启发式文件分类分配策略所改善的性能就越明显.     

On-demand data broadcast scheduling based on AHP and GRA methods in wireless broadcast/UMTS integrated systems

The increasing demand for interactive mobile multimedia service is causing the integration of 3rd generation (3G) cellular systems and wireless broadcast systems. The key challenge is to support data dissemination with low response time, request drop rate, and the unfairness of request drop. This article proposes a novel scheduling algorithm called DAG (on-demand scheduling utilizing analytic hierarchy process (AHP) and grey relational analysis (GRA)), which takes multiple factors-waiting time, number of active requests, deadline-into consideration, and models the data scheduling process as a multiple factors' decision-making and best option-selecting process. The proposed approach comprises two parts. The first part applies AHP to decide the relative weights of multiple decision factors according to user requests, while the second adopts GRA to rank the data item alternatives through the similarity between each option and the ideal option. Simulation results are presented to demonstrate that DAG performs well in the multiple criterions mentioned above.     

An improved hybrid semantic matching algorithm with lexical similarity

In this paper, we proposed an improved hybrid semantic matching algorithm combining Input/Output (I/O) semantic matching with text lexical similarity to overcome the disadvantage that the existing semantic matching algorithms were unable to distinguish those services with the same I/O by only performing I/O based service signature matching in semantic web service discovery techniques. The improved algorithm consists of two steps, the first is logic based I/O concept ontology matching, through which the candidate service set is obtained and the second is the service name matching with lexical similarity against the candidate service set, through which the final precise matching result is concluded. Using Ontology Web Language for Services (OWL-S) test collection, we tested our hybrid algorithm and compared it with OWL-S Matchmaker-X (OWLS-MX), the experimental results have shown that the proposed algorithm could pick out the most suitable advertised service corresponding to user’s request from very similar ones and provide better matching precision and efficiency than OWLS-MX.     

Task scheduling algorithm for system-wide information management based on multiple QoS constraints

An ant colony optimization task scheduling algorithm based on multiple quality of service constraint (QoS-ACO) for SWIM was proposed.Focusing on the multiple quality of service (QoS) requirements for task requests completed in system-wide information management (SWIM),considering the task execution time,security and reliability factors,a new evaluate user satisfaction utility function and system task scheduling model were constructed.Using the QoS total utility evaluation function of SWIM service scheduling to update the pheromone of the ant colony algorithm.The simulation results show that under the same conditions,the QoS-ACO algorithm is better than the traditional Min-Min algorithm and particle swarm optimization (PSO) algorithm in terms of task completion time,security,reliability and quality of service total utility evaluation value,and it can ensure that the user's task scheduling quality of service requirements are met,and can better complete the scheduling tasks of the SWIM.     

R×W: a scheduling approach for large-scale on-demand databroadcast

Broadcast is becoming an increasingly attractive data-dissemination method for large client populations. In order to effectively utilize a broadcast medium for such a service, it is necessary to have efficient on-line scheduling algorithms that can balance individual and overall performance and can scale in terms of data set sizes, client populations, and broadcast bandwidth. We propose an algorithm, called R×W, that provides good performance across all of these criteria and can be tuned to trade off average and worst-case waiting time. Unlike previous work on low overhead scheduling, the algorithm does not use estimates of the access probabilities of items, but rather, it makes scheduling decisions based on the current queue state, allowing it to easily adapt to changes in the intensity and distribution of the workload. We demonstrate the performance advantages of the algorithm under a range of scenarios using a simulation model and present analytical results that describe the intrinsic behavior of the algorithm