摘要
针对WWW业务的非连续特性,提出了一种根据业务参数的变化动态确定移动用户设备信令连接释放时刻的算法.该算法兼顾了会话过程中的信令连接重建概率、信令连接占用效率以及信令负荷三方面因素,网页下载时间和阅读时间的预测采用了具有重尾特性的Pareto模型.此外,算法的输入参数还包括每次会话中网页请求的平均次数,从而进一步增强了优化效果.在信令负荷的预测上同时考虑了状态转移信令和移动性管理信令.仿真结果表明,这一动态算法对信令连接重建概率和信令连接占用效率两方面性能提供了良好的平衡机制,并使控制信令总量维持在一个较低的水平.
Based on the discontinuous characteristic of WWW services, this paper presented an algorithm which dynamically determines the moment of releasing mobile user equipment's signaling connection according to variation of service parameters. The algorithm considers the factors of signaling connection reestablishment ratio, signaling connection occupation efficiency and total signaling overload during the session. Pareto model, which exhibits a heavy-tailed characteristic, is adopted in the prediction of page download time and reading time. In addition, the average number of page requests per session is included in the input parameters of the algorithm, so that the effect of optimization is improved. State transition signaling and mobility management signaling are both considered in the prediction of total signaling overload. The simulation results show that this dynamic algorithm provides a good mechanism of balancing between signaling connection reestablishment ratio and signaling connection occupation efficiency, and keeps a low level of total control signaling quantity.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2007年第5期698-703,共6页
Journal of Shanghai Jiaotong University
基金
国家自然科学基金委创新研究群体基金资助项目(60521002)
科技部科技攻关项目(2005BA908B02)
上海市科学技术发展基金资助项目(05dz05802)
关键词
状态转移
移动性管理
重尾分布
PARETO分布
state transition
mobility management
heavy-tailed distribution
Pareto distribution