Tightly-Coupled Integration of WiFi and MEMS Sensors on Handheld Devices for Indoor Pedestrian Navigation
Tightly-Coupled Integration of WiFi and MEMS Sensors on Handheld Devices for Indoor Pedestrian Navigation
复制标题
DOI：
10.1109/jsen.2015.2477444
复制
发表时间：
2016-01-01
影响因子：
4.3
通讯作者：
El-Sheimy, Naser
中科院分区：
文献类型：
--
作者：
Zhuang, Yuan;El-Sheimy, Naser
在过去的几年里，对室内步行导航仪的需求在各种应用中迅速增加。然而，室内导航仍然面临许多挑战和实际问题，例如需要特殊的硬件设计和复杂的基础设施要求。本文最初提出了一种行人导航的基础上紧耦合（TC）集成的低成本微机电系统（MEMS）传感器和WiFi的手持设备。本文提出了另外两种方法来提高导航性能：1）使用基于行人航位推算/惯性导航系统（PDR/INS）集成的MEMS解决方案; 2）使用运动约束，如非完整约束、零速度更新和零角速率更新的MEMS解决方案。本文有两个主要贡献：1）使用扩展卡尔曼滤波器对WiFi、INS和PDR进行TC融合以进行行人导航，以及2）使用PDR和INS集成以消除仅使用垂直陀螺仪时出现的陀螺仪噪声的更好航向估计。所提出的导航算法的性能已被广泛验证，通过在室内环境中的现场测试。实验结果表明，在三种轨迹下，所提出的TC集成方案的平均均方根位置误差为3.47 m，分别为INS的0.01%、PDR的10.38%、MEMS方案的32.11%和松耦合集成方案的64.58%。所提出的TC组合导航系统可以很好地工作在稀疏部署的WiFi接入点的环境。
The need for indoor pedestrian navigators is quickly increasing in various applications over the last few years. However, indoor navigation still faces many challenges and practical issues, such as the need for special hardware designs and complicated infrastructure requirements. This paper originally proposes a pedestrian navigator based on tightly coupled (TC) integration of low-cost microelectromechanical systems (MEMS) sensors and WiFi for handheld devices. Two other approaches are proposed in this paper to enhance the navigation performance: 1) the use of MEMS solution based on pedestrian dead reckoning/inertial navigation system (PDR/INS) integration and 2) the use of motion constraints, such as non-holonomic constraints, zero velocity update, and zero angular rate update for the MEMS solution. There are two main contributions in this paper: 1) TC fusion of WiFi, INS, and PDR for pedestrian navigation using an extended Kalman filter and 2) better heading estimation using PDR and INS integration to remove the gyro noise that occurs when only vertical gyroscope is used. The performance of the proposed navigation algorithms has been extensively verified through field tests in indoor environments. The experiment results showed that the average root mean square position error of the proposed TC integration solution was 3.47 m in three trajectories, which is 0.01% of INS, 10.38% of PDR, 32.11% of the developed MEMS solution, and 64.58% of the loosely coupled integration. The proposed TC integrated navigation system can work well in the environment with sparse deployment of WiFi access points.