节点文献
异构网络中基于指纹匹配的定位技术研究
Research on Fingerprint Matching Localization Technology in Heterogeneous Networks
【作者】 李虎;
【导师】 罗新龙;
【作者基本信息】 北京邮电大学 , 电子与通信工程, 2015, 硕士
【摘要】 随着无线通信技术的飞速发展,不同的网络环境和用户有更加多样化的业务需求,各种独具特点的无线通信技术应运而生。未来无线通信网络将是由多种无线接入网络相辅相成、相互依存组成的一个统一的异构无线网络。基于位置的服务(Location Based Service, LBS)作为一种新兴的移动增值业务,因其巨大的市场潜力和应用价值,近年来取得了突飞猛进的发展。随着无线网络的融合,LBS在异构网络环境下的发展和应用受到了广泛关注。当前无线接入网络正朝着异构化的方向发展,蜂窝网络与WLAN网络相互重叠的异构场景广泛存在。在异构无线网络中,无线定位技术占据了重要地位。与同构网络相比,异构网络包含有更多的信标节点,能够提供更加丰富的定位观测量,进而提高无线定位的精度。此外,在信标节点分布上,异构无线网络可以克服单一网络由于节点稀疏造成的定位失败和不连续的问题,保证定位服务的连续性和统一性。本文以WLAN/GSM异构网络为研究对象,首先分析了异构无线网络中的信标节点分布情况和室内环境下的接收信号强度特性。针对定位观测量不足和信标节点频繁缺失这两个问题,在课题组研究的基础上,借助蜂窝网络的信号参量增加定位观测量,提出了一种有效的基于代价函数的指纹匹配定位算法,并在实际环境中验证了算法性能。最后,基于轨迹信息的连续性,构建空间上下文约束区域,对所提出的指纹匹配定位算法做了进一步优化。最后,从实时定位跟踪的需求出发,设计和实现了异构无线网络下的实时定位跟踪系统。基于Android平台和百度地图SDK开发实现了用于实时定位和位置显示的移动客户端应用,基于高性能网络服务器和数据库开发实现了定位服务中心,为多用户实时定位提供后台支撑。
【Abstract】 In recent years, a variety of wireless communication technologies with different features arise due to different network environments and service requirements. Future wireless networks will be heterogengeous networks that consist of a variety of wireless access technologies. In addition, location based service (LBS) is a new value-added service and has achieved great development because of its huge market application value. With the fusion of wireless access networks, LBS in heterogengeous networks has been widely concerned.Heterogeneous networks consisting of cellular networks and WLAN networks are widespread in real wireless environments. In such heterogeneous environments, wireless localization technology plays an important role in many applications. Compared with homogeneous network, more RF signal features and more uniform distribution of anchor nodes in heterogeneous networks can be applied in fingerprint matching technologies to provide an accurate, reliable and continuous location service.This paper studies the fingerprint matching localization technologies in GSM-WLAN heterogeneous networks. The fingerprint matching localization problem in sparse WLAN areas is focused. Multiple RSSI values from different networks are used to rich the location information of the measurement vector and a fingerprint matching localization algorithm based on cost function is proposed to reduce the impact of missing anchor nodes. In order to filter the singular location fingerprints, a spatial-context constraint area is constructed based on the continuity of trajectory to improve the localization accuracy. Finally, from the demand of real-time positioning and tracking, a real-time positioning and tracking system based on heterogeneous wireless networks is designed and implemented. A mobile localization client application is developed based on Android platform and Baidu map SDK to locate and display the current position of user terminal and a localization server is implemented based on high-performance network server and SQL database to provide support for real-time localization service.
【Key words】 Heterogeneous Networks; Wireless Localization; Fingerprint Matching; Cost Function; System Implementation;