多排队多服务台的服务系统病床安排优化研究

采用运筹学排队论原理对某眼科医院门诊病人的排队系统进行研究,本文提出各种眼科疾病病床设置的最佳值,并与医院现在的住院安排做对比,为医院的管理提供改进建议和科学的理论依据。该方法可以应用于其他疾病的病床安排。将该论理论应用于医院的病床安排是科学、有效的方法。     

带准备时间的柔性流水车间多序列有限缓冲区排产优化问题

针对带准备时间的柔性流水车间多序列有限缓冲区排产优化问题,提出一种改进的紧致遗传算法（Improved compactgenetic algorithm,ICGA）与局部指派规则结合的方法来解决该问题。全局优化过程采用改进的紧致遗传算法,为了克服紧致遗传算法（Compact genetic algorithm,CGA）易早熟收敛的问题,提出一种基于高斯映射的概率模型更新方式,在保持紧致遗传算法快速收敛特性的前提下,扩展了种群中个体的多样性,增强了算法进化活力。为减少生产阻塞和降低准备时间对排产过程的影响,设计了多种局部启发式规则来指导工件进出多序列有限缓冲区的分配和选择过程。采用某客车制造企业中的实例数据进行测试,测试结果表明,改进的紧致遗传算法与局部指派规则配合使用,能够有效解决带准备时间的柔性流水车间多序列有限缓冲区排产优化问题。     

基于多队列和多线程的短信实时并发控制算法

针对Modem控制的短信应用在实时性和并发性等性能上的不足,以实际项目为背景,提出基于多队列和多线程的短信实时并发控制算法。以排队论模型为依据,采用多队列、多线程、池等技术手段保证短信控制的实时性、并发性以及数据的共享性。通过排队论模型和实验对算法进行验证,结果表明,该算法大幅提高了短信发送和接收的速度与效率,满足了短信应用的实时性和并发性需要,优化了短信应用系统的整体性能。     

Analytical modeling of a multi-queue nodes network router

This paper presents the derivation of an analytical model for a multi-queue nodes network router, which is referred to as the multi-queue nodes (mQN) model. In this model, expressions are derived to calculate two performance metrics, namely, the queue node and system utilization factors. In order to demonstrate the flexibility and effectiveness of the mQN model in analyzing the performance of an mQN network router, two scenarios are performed. These scenarios investigated the variation of queue nodes and system utilization factors against queue nodes dropping probability for various system sizes and packets arrival routing probabilities. The performed scenarios demonstrated that the mQN analytical model is more flexible and effective when compared with experimental tests and computer simulations in assessing the performance of an mQN network router.     

Load balancing geographic routing strategy for aeronautical ad hoc networks

In aeronautical ad hoc networks, the traditional greedy perimeter stateless routing (GPSR) protocol poses sev-eral issues. For example, it is difficult to adapt to the highly-dynamic network environment, and it is prone to cause con-gestions. In order to address the problems, a TTE (time to enter the communication range of the destination)-based load balancing geographic routing (LBGR) protocol was presented. Taking TTE as the main routing decision metrics, this pro-tocol included the TTE-based packet forwarding scheme, multi-path traffic allocation scheme, and local optimum han-dling scheme. Furthermore, the multi-path traffic allocation scheme employing the queueing theory was modeled, and the mathematical expressions of some metrics were derived, such as the mean queue size, mean number of packets waiting in the queue, and mean waiting time. Finally, the analysis of the OMNeT++ simulations shows LBGR protocol has advan-tages over GPSR and some other protocols in terms of the packet delivery ratio and end-to-end delay, and is more suitable for the highly-dynamic aeronautical environment.     

基于事件驱动的Unix并发控制算法

以实际项目为开发背景,提出了基于事件驱动的并发控制算法,采用多线程、多队列和线程池等技术来保证系统的实时性、并发性和数据共享性。通过仿真对该算法进行了验证,结果表明,该算法能够实现系统的功能,提高系统的运行效率。     

多重分形网络流量输入下多队列排队模型

针对目前分形网络流量输入下排队系统研究大多没有给出解析结果,且只研究单队列排队的缺点。根据实际网络的排队性能,在无约束条件下,采用了多队列多服务台成批到达的排队方式,建立模型,推导出多重分形特性网络流量输入下单队列和多队列溢出概率的解析式。并通过真实网络流量数据描绘出溢出概率与队列缓存长度的关系,结果发现该多队列排队模型下的溢出概率的解析式能够较好地拟合真实网络流量的排队性能。     

EPON系统DBA门限策略的数学建模与分析

研究了以太无源光网络（EPON)系统服务门限的选取策略,提出了一种扩展的单服务台多终端站轮询模型（EL-CSMQ）. 该模型首先在输入侧建立符合自相似特性的批到达流量模型,随后采用两级缓存的建模方法,分析了EPON系统动态带宽分配(DBA）的门限大小对系统稳定性、平均时延和时延抖动的影响. 仿真验证了理论分析结论.     

交叉流水线多队列缓冲报文转发技术

IP报文封装为链路帧是路由器设计必不可少的技术。提出了一种通用的多通道报文封装和转发的处理器结构．利用FPGA内部存储资源,采用流水线和多队列缓存区相结合,显著提高了小报文线速转发和突发流量传输的性能。