Heading Misalignment Estimation Between Portable Devices and Pedestrians

A. Ali, H-W. Chang, J. Georgy, Z. Syed, C. Goodall

Abstract:	The last few years witnessed major advancements for mobile phones in different respects that enabled them to be very capable devices. These phones can be used for positioning and navigation for different application such as Location based services (LBS) and portable navigation and guidance. A major source of positioning in smartphones is the Global Navigation Satellite Systems (GNSS) receivers they have onboard. Smartphones come equipped with Assisted Global Positioning System (AGPS) chipsets, and recently some receivers in the phones are capable of using other constellations such as GLONASS. Nowadays, smartphones also include sensors, such as accelerometers, gyroscopes, magnetometers, and barometers. Currently, Micro Electro-Mechanical System (MEMS) sensors are used predominantly for entertainment applications in smartphones. Furthermore, the orientation of the screen is sensed by the MEMS accelerometer which is also used to switch the display according to the user's needs. There are some applications that use the accelerometer and magnetometer to provide a navigation solution if the user is careful enough to keep the device in a specific orientation with respect to their body, such as when held carefully in front of the user after calibrating the magnetometer; such careful holding is not very realistic. Some recent research works are including gyroscopes for navigation. In general, the embedded mobile phone sensors are considered insufficient for reliable navigation purposes due to very high noise, large random drift rates, and especially for a mobile device that freely changes orientation with respect to the carrying platform or person. In traditional inertial navigation, alignment of the inertial sensors within the platform and with the platform's forward, transversal and vertical axis, is critical. If the inertial sensors, such as accelerometers and gyroscopes are not exactly aligned with the platform, the positions and attitude calculated using the readings of the inertial sensors will be far from optimal. Fixing the inertial sensors within the platform is thus a requirement for navigation systems that provide high accuracy navigation solutions. For tethered systems, one known mean for ensuring optimal navigation solutions is to utilize careful manual mounting of the inertial sensors within the platform. However, portable navigation devices or navigation-capable devices such as smartphones, are able to move whether constrained or unconstrained within the platform which can be a person or a vehicle, so careful mounting is not an option. There is a need for a navigation solution capable of accurately utilizing measurements from a portable device being held, carried or used by a person moving on foot, where this solution is able to determine the navigation state of the device and the person carrying the device without any constraints on the person to move indoor/outdoor, or how the person uses the device. The estimation of the position and attitude of the person has to be independent of the usage of the device, e.g., the way the user is holding or moving the device during navigation. This highlights the key importance of misalignment determination between the device and the person, to enable the portable navigation device to be used in any orientation. Some techniques available in the literature are able to calculate just discrete or pre-determined values of the heading misalignment angle based on discrete use case classification of the device. This limits their usage to these discrete use cases; and even when one of the supported use cases is used, the accuracy can deteriorate if the true misalignment value is somewhat different then the discrete misalignment value assigned with the classified use case. The proposed technique is able to get a heading misalignment angle, i.e., the angle between the device and person headings, covering the whole misalignment space, not just discrete or pre-determined values of such angle. The presented technique uses the inertial sensors’ readings to determine the misalignment angle, and it can be used to give an output at a rate equal to or less than the rate of the inertial sensors’ readings. The proposed technique can work whether in the presence or in the absence of absolute navigation sources such as, for example, GNSS or WiFi positioning. If absolute navigation sources are temporarily available, they can help enhancing the misalignment estimation; also this enhancement can take effect in periods of blockage or unavailability of absolute navigation information after they were available. The technique presented in this paper can work with various types of walkers with different gaits and speeds. The proposed technique can also give an accuracy measure for the quality of the misalignment estimation. The proposed technique was tested on a very large amount of trajectories from a lot of people with different walking types, gaits, and speeds. Following are some scenarios that were part of the testing: 1) start outdoors in parking lot and enter building with minimal multipath, 2) start outdoors in parking lot and enter building with severe multipath, 3) start indoor from a known position and navigate without any GNSS for the whole test trajectory, 4) start the navigation and stay in a severe multipath environment (downtown core). Scenarios 1 to 4 were conducted with diverse set of orientations covering the different uses cases of smartphones and portable devices including but not limited to handheld in different orientations and tilts, hand dangling with different orientations and tilts, ear mode with different tilts, belt with different orientations and tilts, pocket with different orientations and tilts, chest or back with different orientations and tilts. Scenarios 1 to 4 were repeated where the trajectory contained a single use case and they were also repeated for trajectories that contain multiple varying use cases. The results were assessed; plots of the device roll, pitch, and heading, as well as the estimated misalignment, and the person heading were studied, and finally the positioning results were plotted on digital maps to analyze the misalignment estimation effects on positioning performance. The results clearly demonstrated the efficacy of the proposed technique and how it can enable a real-time, continuous and reliable consumer localization indoors and outdoors with mobile devices.
Published in:	Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013) September 16 - 20, 2013 Nashville Convention Center, Nashville, Tennessee Nashville, TN
Pages:	1626 - 1633
Cite this article:	Ali, A., Chang, H-W., Georgy, J., Syed, Z., Goodall, C., "Heading Misalignment Estimation Between Portable Devices and Pedestrians," Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013), Nashville, TN, September 2013, pp. 1626-1633.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In