一种方向性预测的位置管理算法  

Algorithm improvement of directional location management

作　　者：陈杰[1,2] 奚文 莫玮[1] 

机构地区：[1]西安电子科技大学机电工程学院,陕西西安710071 [2]长江计算机(集团)公司,上海200001

出　　处：《西安电子科技大学学报》2015年第3期83-89,共7页Journal of Xidian University

基　　金：软件和集成电路产业发展专项资金资助项目(RX-RJJC-03-10-3803)

摘　　要：为了减少无线网络位置的管理开销,提出一种新的采取方向性预测的马氏链作为移动模型.在二维六边形蜂窝网络结构下,建立两个连续呼叫时间间隔内的位置更新模型.推导计算出呼叫到达和小区驻留时间分别服从泊松分布和指数分布时,移动台移动距离的最佳门限值及门限值范围内穿越K个小区边界的概率公式.并基于方向预测的概率分布给出了最优两步寻呼的改进算法策略.最后,通过数值仿真分析,表明所给出的寻呼策略比其他"基本策略"更优.In order to accurately calculate the cost of location management for the wireless cellular network, this paper puts forwards the predictive directional Markov chain as a mobility model based on the two-dimensional hexagon cellular structures. According to the location update model based on the two-steps paging strategy established between two consecutive arrival calls to get the optimal threshold of movement-based location management and the probability distribution of crossing the cell boundary K times on the condition that the incoming calls form a Poisson process and the MT＇s cell residence time has exponential probability distribution. Then with these given probability distribution we give the optimal two-steps paging strategy. Finally, numerical results show that this strategy outperforms some well-known basic paging schemes.

关 键 词：位置管理 位置更新 两步寻呼 开销分析 

分 类 号：TN929.53[电子电信—通信与信息系统]