基于AGA优化BP网络的小信息量蜂窝网定位算法.pdfVIP

第 10 卷第 4期 信 息 工程 大学 学 报 V o l10 N o4 2009 年 12 月 Journal o f Inform ation Eng ineering U n iversity D ec2009 AGA BP 周 艳, 王建辉, 胡捍英 ( , 450002) : 为了实现蜂窝网中小信息量情况下对移动用户的定位, 在只有驻留基站和 台定位设备 可用的条件下, 建立了TOA /TOA 定位模型, 提出了 种基于BP 神经网络的定位算法, 并用自 适应遗传算法(A GA )优化 BP 网络的权值仿真结果表明, 在小信息量条件下, 优化后的算法 在复杂的多径环境下有效地提高了定位精度, 能够很好地解决模型中固有的模糊点问题, 达到 了FCC 的定位要求 : BP 神经网络 AGA 移动台 定位 : TN 9295 3 : A : 1671- 0673( 2009) 04- 0461- 05 Cellu lar Location A lgorithm w ith Less Inform ation Based on Op tim ized BP N eu ral N etw ork by AGA ZHOU Y an, W ANG J ianhu,i HU H anying ( In stitute o f In fo rm ation Eng ineer ing, Inform ation Eng inee ring U niversity, Zheng zhou 4 50002, Ch ina) Ab stract In order to rea lize m obile location w hen less param eter inform ation can be obtained in cel lu lar network s, this pap er presents a location m odel b ased on TOA /TOA when on ly the b ase stat ion and one location dev ice are available. T hen an a lgorithm b ased on BP ( Back P rop agation) is intro du ced to realize position estmi ation under th is m odel, and the w e ights o f BP newt ork is optmi ized by A GA ( adaptive genetic algorithm ). Smi u lation results indicate th at the location accuracy of U E ( u s er equ ipm ent) is mi proved effectively in com plex m ultip ath env ironm en t and the blurred spot prob lem th at inherently ex ists in th is m odel can be so lved by the optmi ized BP algo rithm. The perform ance o f the a lgorithm m eets the dem ands of FCC. Key w ord s BP neural netwo rk AGA m obile locat ion less inform ation TOAAOA , ,