【作者】 王爱民；

【导师】 苑森淼；

【作者基本信息】 吉林大学 ， 通信与信息系统， 2004， 博士

【摘要】 移动通信是20世纪经济、能源、交通运输与通信技术高速发展相结合的产物。在20世纪的最后十年，移动通信特别是数字移动通信发展之快和应用之广，大大超出了人们的预料和专家的预测。移动通信的发展经历了模拟和数字两大历程，数字移动通信体系中，目前应用最多的是时分多址（Time Division Multiple Access，缩写为TDMA）和码分多址（Code Division Multiple Access，缩写为CDMA）两种技术。CDMA技术比TDMA技术投入商用较晚，却显示出强大的生命力。CDMA技术代表着第三代移动通信。论文介绍了移动通信发展的历史，总结了CDMA网络的体系结构组成；对于CDMA网络中的各项关键技术进行了描述；总结了CDMA网络需要深入讨论的一些问题，包括无线信道资源管理、软切换策略、无线接入研究、功率控制和负荷控制等；对于CDMA网络QoS研究的背景进行了陈述。介绍了现有QoS管理的各种机制，并通过对CDMA网络标准中QoS的各种规定的回顾，分析了QoS管理机制在CDMA网络中的应用。在简介已提出的一些CDMA网络接入模型的基础上，充分考虑了CDMA网络的特点，提出了新的预测模型。依据预测模型来测量负载的变化，并针对这些变化调整QoS参数，期望获得较好的QoS性能，并在系统繁忙情况下可以适当牺牲QoS性能来换取系统的正常运行。这一策略强烈依赖预测模型的计算结果，但是因为离散的采样数据结果不能直接参与连续模型的计算，需要插值处理，这增加了系统的计算代价，而且造成计算的不准确。为此进一步提出了离散化的模型，从而避免了模型用插值方法迭代求解的困难，对于简单和复杂的运行环境具有很好的自适应能力，同时在处理突发接入请求方面具有较好的效果。 介绍了CDMA网络中软切换所依赖的各类参数特性，进一步阐述了各个参数在软切换中的作用。软切换过程中经常出现的移动台速度突变、移动台乒乓移动以及速度过于缓慢的移动台造成小区拖尾现象，这些因素容易造成软切换失败，产生移动台通信过程中掉线的不良效果。传统的软切换算法在处理这些问题的时候基本上是以牺牲容量来换取宏分集成功的策略，即让一个移动台占用几个信道来避免软切换的失败。这在系统空闲时不失为一个好的方法。但是系统繁忙时上述条件不能被满足。 <WP=113>为此，首先描述了单队列软切换算法，以优先级标记的方式来管理提出切换要求的移动台导引信号，可以有效的识别虚假切换。进一步针对单队列算法对于虚假切换的激烈反应缺陷描述了双队列算法，双队列算法动态的管理虚假切换，在必要的时候可以将虚假切换恢复成真正的切换，保证了切换的正常进行。队列算法解决了虚假切换的问题，但是应该进一步提出有效的办法解决移动台速度突变、乒乓移动、小区拖尾等问题。首先提出了动态门限值产生办法，门限值依赖宏分集、切换紧迫程度动态产生，使得宏分集和切换之间维持了一个平衡，既保证了软切换成功进行又保证了系统的容量。在动态门限值的基础上，为了解决速度突变问题，设置一个队列来存放导引信号强度变化率，并动态排序，基于统计选取测量窗口后确定调整因子，将其作用于动态门限值，在不改变切换策略的前提下完成了对于速度突变问题的处理。通过考察导引信号强度变化率，进一步考虑数学符号的变化，并进行了离散化、等距离处理，运用其累计效果描述移动台的变化，使得系统能够识别乒乓移动的移动台，避免了该类移动台对于系统资源的占用以及反复响应所引起的计算代价。小区拖尾是因为移动台速度过于缓慢，导致其导引信号强度变化率小到几乎被忽略。系统基于统计策略得到一个适当累计值，去除符号后，该累计效果通过放大效应，可以准确描述缓慢移动台的请求紧迫度。多个服务小区对移动台进行宏分集后，对于宏分集的导引信号强度进行排队，产生动态切换门限值，对于低于该门限值的导引信号，实行切换，保证系统的容量。该方法避免了传统固定门限值的僵化，计算代价小。对于提出的各类算法，给出了一个平衡模型，在模型的统一协调下，各算法不再是孤立追求QoS效果的节点，而是服务于整个系统，追求统一QoS效果的支撑体系。为了验证各算法的实际应用效果，进行了一系列试验。试验表明各算法应用效果明显，能有效的消除软切换过程中坏条件应用下对于系统的影响。为了模拟特定的坏应用条件，提出了借助于交通流分布模型的方法，为了适应CDMA网络条件进一步进行了泊淞改造，并做了适当的离散化，以适应试验算法的输入。如果进一步结合QoS分层模型讨论软切换问题将使得算法的应用环境更加具有针对性。因为信道混合分配策略HCA的性能和固定信道与动态信道的配置比例有直接关系，即不同的配置比例会导致不同的系统容量，所以选择合适的比例至关重要。本文在分析固定信道容量和动态信道容量的基础上，提出了该比例的产生规则，并给出了定量的制约计算方程。试验表明在该方程的约束下进行信道分配，应用HCA策略的整体容量上升，可以得到较好的QoS效果。另外给<WP=114>出了HCA策略下的系统预警方法。 不同QoS要求等级的多类综合业务同时依赖于CDMA网络进行传输，共同形成对于信道资源的竞争。本文基于不更多还原

【Abstract】 Mobile communication is the result of connect among development of economy and technology and communication. In last decade of 20th century, mobile communication especially digital mobile communication developed unexpectedly faster and applied more widely. Mobile communication live two phase: simulation communication and digital communication. In digital mobile communication system , the most wide use of application is TDMA(Time Division Multiple Access) and CDMA(Code Division Multiple Access). CDMA is later than TDMA in business, but it appears good vitality. CDMA is representation of the 3G mobile communication system. The paper introduces the development history of mobile communication technology and gives summary of the compose for CDMA system structure, then describes each pivotal technology in CDMA system. The paper summarize several questions in CDMA system which needs to research more deeply and includes the management of wireless channel resource, the policy for soft handoff, research in wireless access, control of power and control of load etc. Then I mention the background of QoS(Quality of Service ) research in CDMA network which I will discuss about. Through review every rules in CDMA network’s QoS standard, the paper introduce all kinds of mechanism in QoS management and analysis the QoS management of application in CDMA network. Based on some access model in CDMA network which I mentioned , I provides a new forecast model to forecast the load change, depend on the change the CDMA system adjust the QoS parameters in order to gain better performance in QoS management. Another when system is busy, the system can depress whole QoS performance properly to ensure that system can run placidly. The policy rely on the result of forecast model strongly. But the result of sampling data is discrete and the continuous model can not calculate directly on it. So we must give interplation conversion result in the increase of calculate cost and made imprecise calculate result. In order to resolve this problem I provide this new discrete model. It can avoid the difficulty of calculating based on interplation and it has good self-adapt ability in simple or complex environment. At the same time, it can obtain good effect in deal with paroxysmal access request. I also introduce each parameter’s characteristic that can impact the soft <WP=117>handoff QoS feature in CDMA network and then describe it’s use. During soft handoff process, some questions are often occur, such as the paroxysmal speed change to mobile station and it’s ping-pong moving and tail drag in cell because of slow speed. It can lead soft handoff lost and outage during wireless communication. Traditional soft handoff arithmetic choose a policy to exchange the success of macrodiversity through depressing the capability of the whole system. That is a mobile station occupy several channel to avoid the fail on soft handoff. When system is not busy, this policy is a good method. But when system is busy all of these conditions can not be met. The paper describe an arithmetic in the condition of single queue at first, it manage the soft handoff request through give priority mark to the that the guidance signal that mobile station bring on. It identify the illusive handoff request effectively. And then I describe an double queue arithmetic that in order to avoid the bug of scorching response with single queue arithmetic on illusive handoff. This method manage illusive handoff dynamically, and can change illusive handoff request to real handoff request when the system need and ensure the soft handoff process can be run correctly. The queue arithmetic solve the illusive question but should solve the other question ulteriorly such as mentions above. The paper give a definiens of dynamic threshold value. The threshold value can be made dynamically depend on exigency level of macrodiversity and soft handoff. It make balance between macrodiversity and soft handoff, not only ensure the soft handoff succ更多还原