中文摘要：

无论是普适计算还是物联网,实时定位均是其关键技术之一。在建筑物、洞室等复杂环境中,由于无线信号的不规则衰减、非视距传播,使得常用的一些定位算法（如接收信号强度、到达角度等）无法达到较高的定位精度。为了解决复杂环境中定位精度不足的问题,设计并实现了一个定位精度改进中间件。它集成了平滑、传感器融合、卡尔曼滤波、粒子滤波等多种精度改进策略,可以根据用户对定位精度的需求动态地选择相应的策略。通过实验分析了不同精度改进策略的性能,结果表明,该中间件能够较好地满足用户对不同精度的需求。

英文摘要：

Real time positioning is an essential technology for both ubiquitous computing and Internet of Things.In a building,cavern or other complex environments,due to irregular attenuations and non-sight transmissions of wireless signals,some common positioning algorithms such as received signal strength,angle of arrival etc.,fail to achieve high positioning accuracy.To solve the inaccuracy problem in a complex environment,a positioning precision improvement middleware is designed and implemented.It integrates smoothing,sensor fusion,Kalman filter,particle filter as well as other precision improvement strategies so that it allows users to choose a corresponding strategy according to positioning accuracy requirements.The authors also analyzed through experiments the performances of different precision improving strategies.The results showed that the middleware can meet various accuracy requirements from users.