Robust Integrated Indoor Navigation Using Consumer Device Sensors, Map Matching and Wireless Positioning

Tao Li, Jacques Georgy, Zanaib Syed and Chris Goodall

Abstract:	Seamless outdoor and indoor positioning using portable devices has become an important requirement for many emerging applications. Therefore, Global Navigation Satellite Systems (GNSS) receivers, low cost Micro Electro-Mechanical Systems (MEMS) inertial sensors, barometers and magnetometers have been the standard configuration for many current smartphones, smart watches and tablets to provide positioning information. Although Pedestrian Dead Reckoning (PDR) based portable navigation system with GNSS and multiple sensors is able to provide promising positioning results in most outdoor environments, in the absence of GNSS signals in indoor or deep urban environments, it still lacks robustness because of the accumulated heading error and step length error, in addition to the challenges of portability and free change of the orientation and usage of the device by the user. These shortcomings will cause a skewed path over time and produce position estimates that might not be consistent with the building layout. The resulting paths cross walls, ?oors or other obstacles. Although magnetometer measurements can be used to update the heading, they are vulnerable to interferences which are quite common in complex indoor environments. This paper proposes a robust indoor navigation solution integrating low-cost sensors in consumer devices with indoor map matching and wireless positioning. The user can operate the portable device freely without constraints, such as talking on the phone, dangling the phone beside them, texting, putting the phone in their pocket, on their belt or in their backpack. The motion context detection algorithm can detect the user’s motion patterns such as walking, fidgeting, going up/down stairs, standing/walking on escalators, or using elevators. The proposed map matching engine is based on multi-hypothesis Kalman filter and is very computationally efficient. The map match engine can manage the hypothesis creation, trim and removal based on the user position uncertain and the motion context. These motion contexts can be used to apply 2D position update to improve positioning performance and also to initiate floor changes. Furthermore, any incorrect misalignment between portable devices and the human body can be monitored and corrected based on the current indoor environments. Moreover, the map match engine analyzes the history of the user’s trajectory and indoor environments to improve the step length estimates from PDR. Finally, the indoor map is combined with WiFi positioning to derive a more reliable weighting scheme based on the indoor environments. Also, with the map information, WiFi blunders can be more reliably removed. The proposed system can also use heading and speed estimated from the wireless system, if they are available; it is also capable of taking other wireless updates if they are available, such as for example from Bluetooth low energy (BLE). Real-time test results are presented to demonstrate the performance of the proposed portable positioning system.
Published in:	Proceedings of the 28th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2015) September 14 - 18, 2015 Tampa Convention Center Tampa, Florida
Pages:	15 - 23
Cite this article:	Li, Tao, Georgy, Jacques, Syed, Zanaib, Goodall, Chris, "Robust Integrated Indoor Navigation Using Consumer Device Sensors, Map Matching and Wireless Positioning," Proceedings of the 28th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2015), Tampa, Florida, September 2015, pp. 15-23.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In