【作者】 王芙蓉；

【导师】 黄载禄；

【作者基本信息】 华中科技大学 ， 通信与信息系统， 2004， 博士

【摘要】 移动通信是当代通信领域内发展最迅速、潜力最大、市场前景最广、最引人瞩目的热点技术,也是最活跃的领域之一。移动网的重要特点在于用户的移动性,因此它要求网络具有移动性管理功能。 立足于采用灵活的位置管理网络结构,充分考虑适应于移动用户的个性化特征,本文针对网络侧的位置数据库和接入侧的位置更新和寻呼中的一些关键问题,进行了系统深入的研究,提出了一系列算法和策略,并通过建立仿真模型和理论分析模型证明本文提出的方案有效的提升了系统性能: 搭建了一个基于离散事件系统仿真的综合移动管理仿真环境。描述并抽象了仿真模型中的主要功能化模块,提出了仿真环境的整体结构及主要模型的仿真实现方法。在拓扑模型中提出并完整定义了一种基于方向的蜂窝编码方案,能够反映移动用户运动的方向性,为适应于用户运动特性的位置管理方案的研究奠定了基础。 提出了一套基于移动用户个体运动特性的自适应动态位置管理策略,即基于动态距离更新的阈值计算方法和基于扇形覆盖的方向跟踪位置区定义方法。在动态距离更新策略中,通过对用户个体运动方向的判断,提出了两步寻呼策略,比较得到两步寻呼策略的寻呼代价低于一步寻呼策略的结论。通过仿真比较,可以得到在个体移动用户有一定规律的移动下,本文提出的基于扇形覆盖的方向跟踪位置区定义方法降低了移动台的更新代价。 针对移动通信蜂窝网固定位置区移动性管理中存在的两个主要问题:边界乒乓效应和边界信令集中,本文提出了一种边界关联更新策略。当移动台跨越位置区边界,立即引发位置更新时,将该移动台在原位置区的“拖尾”蜂窝作为新位置区的扩展,从而在很大程度上抑制了乒乓效应,同时,部分平滑了位置区边界的信令负荷。采用随机游走模型和状态转移矩阵,对本文更新策略与传统更新策略进行了比较,证明本文提出的边界关联更新策略将更新代价降低了约 25%,并通过实际仿真进一步验证了结果。分析和实验表明,本文策略不仅相对易于实现,而且极大优化了性能。 以下一代移动网支持多模通信为基础,利用多跳蜂窝网中各移动台之间可对等通信的手段,本文提出一种成组进行位置管理的策略。分析了彼此临近的移动台组成一个自组织网的基本网络模型,提出了组管理策略,即移动用户向组代表注册,由组代表进行信令压缩,构建组控制信令,成批向网络报告位置的位置更新策略。采用M/M/n/n 模型具体比较了组管理方案与传统方案的性能,说明该组方案不仅减少了移动台的发射功率,还大量节省了频率资源,增加了系统的吞吐能力。本文还具体分析 I<WP=4>了成组进行位置管理的时延问题,给出了在时延约束下最佳的组规模计算方法。在分析研究位置注册和呼叫传递的相关研究工作基础上,本文给出了前向指针与局部锚点的改进思路。通过定义 VLR 不同的工作模式,描述了一种关联用户的层次位置管理数据库方案。使用地理上广泛分布的 VLR 作为中央用户数据库基于移动用户的异地备份,提出了一种分布式复制位置数据库方案,并具体分析了其工作过程。最后,将上述思想结合起来,给出了一种混合策略的分布式复制位置数据库方案。论文的研究工作基本覆盖了位置管理研究领域的主要内容,具有一定的理论和实际意义。更多还原

【Abstract】 Mobile communication technologies have experienced tremendous growth over the pasttwo decades. The third-generation systems and the beyond third-generation systems providevarious types of services such as voice, circuit/packet data, and multimedia. The growingdensity of subscribers and the need for hand-off and roaming underscore the importance oflocation management. This dissertation focuses on location management in mobile network, covers locationupdate strategy, paging policy and location database design. Several novel strategies havebeen proposed and verified to be effective by both analysis and simulation. A framework forsimulation system is presented as well. To minimum location update and paging costs, this dissertation proposes a dynamicdistance-based location update scheme, including dynamic location size calculation and atwo-step paging policy. Considering about mobile uses’ mobility characteristics, asector-based dynamic location area model is presented. Simulation result shows thissector-based dynamic location area model has effectively reduced location update costs. In the 4th section, we propose a novel strategy of correlation boundaries against the twomain problems of location management scheme in cellular networks: Ping-Pong locationupdate effect existing between two location areas and concentration of signals at boundaries.This strategy is able to reduce the Ping-Pong effect effectively and smooth the load ofsignals at boundaries by adding certain cells as expansions of the new location area,whenever a mobile terminal moves into a cell belongs to a new location area. Comparedwith existing strategies, the strategy in this paper is easier to implement and also improvesthe performance greatly. Analysis and simulation results both show the correlationboundaries strategy reduces location update costs by about 25%. In 3G and beyond 3G mobile systems, there is a tendency of combination ofheterogeneous networks, such as traditional mobile network, WLAN, and Ad Hoc. Based onthis idea, this dissertation proposes a new location update scheme for subscribers on an areawhere a large number of fast-moving users need send update information frequently. Thegroup location update scheme in which a group location update is performed by a leadermobile terminal instead of individual update by every normal mobile terminal has been III<WP=6>proposed. Mobile terminals (MT) relatively near form a group and select a leader to collectlocation update reports of other MTs and report in groups. We build a stochastic servicesystem model to analyze the group location update scheme and individual location updatescheme in the view of channel access cost. Compared with individual scheme, group updatescheme obviously increases channel capability and system throughput due to a significantreduction in signaling cost per update and the number of location updates. As user mobilityincreases or the number of users in the group increases, the advantage of the group updatescheme improves. We also analyze delay caused by bulk process and give arecommendation for the bulk size choice to get better throughput performance within anacceptable delay. This dissertation’s research work also includes location database design in the networkside. An improved scheme for forwarding pointer and local anchor is presented. Using VLRdistributed in different locations as remote backup for central user profile database, thisscheme approaches a distributed copy location database and by defining different workmodes it can describe related users hierarchical location management database.更多还原