Topic: Alternatives to Kalman Filters
Speaker 1: Christian Phillips
Title: A Deep Learning Approach for the Classification of Multipath Ranging Errors in Challenging Urban Environments
Abstract: Distortion to the correlation function caused by multipath and non-line-of-sight signals can result in pseudorange errors on the order of several tens of meters in urban canyon environments. To address this problem, a deep learning approach for classifying multipath ranging error from a global navigation satellite systems (GNSS) receiver correlation function is presented. This approach uses a one-dimensional convolutional neural network, suitable for embedded applications, to classify the magnitude of pseudorange error associated with correlation functions. The network is trained and tested on live GNSS data collected in a challenging urban environment, and the capability of the model to remove high error measurements for a least-squares positioning solution is explored. The network has proven to be effective at detecting measurements with high multipath ranging error, and the removal of detected measurements reduced positioning error by up to 80%.
Bio: Christian Phillips is a graduate student in the Department of Geomatics Engineering at the University of Calgary. His research focuses on leveraging artificial intelligence to improve the performance of GNSS receivers in challenging operational environments. He is also a Software Developer at Hexagon’s Autonomy and Positioning division. He received his B.Sc. degree from the University of Manitoba in 2022.
Speaker 2: Ilyar Asl Sabbaghian Hokmabadi
Title: Computationally Efficient Particle Filtering for Fusing Angle of Arrival Beacons and IMU Measurements in Indoor Localization Applications
Abstract: In the recent past, beacons have emerged as a promising technology that meets the accuracy and reliability requirements of indoor localization. Due to the challenges regarding the loss of line-of-sight, indoor beacons often cannot provide a consistent performance throughout the navigation in standalone mode. Thus, the fusion of beacons with other sensors, such as inertial measurement units (IMU), has become an important topic for researchers. In recent decades, many estimation techniques have been proposed to achieve such sensor fusion. Among these, the Kalman filter family of estimators are ubiquitous due to their low computational cost. However, these classic estimators require an assumption of Gaussian distribution for the state variables (e.g., position, velocity, and attitude). Unfortunately, this simplistic assumption is not met in real-life scenarios. This research proposes an alternative approach based on particle filtering to fuse angle of arrival (AoA) beacon observations and inertial measurements. First, the theoretical background for reducing the dimensionality of state variables using AoA beacons is shown. This dimensionality reduction will contribute to reducing the computational cost of the particle filter. Second, it is shown that a low-cost and cm-level positioning can be achieved using only two beacons with the help of the proposed particle filter.
Bio: Ilyar Asl Sabbaghian Hokmabadi received his M.Sc. in Geomatics Engineering from the University of Calgary in 2018. Later, he received his Ph.D. degree from the University of Calgary in 2023. During his Ph.D., he developed many localization solutions using mobile and handheld systems. He has published and contributed to different areas, including indoor mapping using ultrasonic sensors, accurate 3D reconstruction using monocular cameras, and multisensory positioning solutions in indoor environments. Currently, Ilyar is an algorithm designer at Profound Positioning Inc., where his responsibilities include exploring state-of-the-art deep learning and advanced optimization methods to achieve a system-wide calibration of different types of sensors.
Location: Room 207 – Engineering Block G (ENG), University of Calgary Campus
Date: Thursday November 28, 2024
Time: Doors will open at 11:30am, presentation beginning at noon
Cost: $20 non-members, $18 members, $15 grad students, undergrad students $10, includes a light lunch and refreshments. All proceeds go towards two annual scholarships for students attending the University of Calgary