Improving the performance of Wi-Fi indoor localization in both dense and unknown environments

• Main Author: HAI, Quang Truong
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2024
• Subjects: Indoor localization; Dense environment; Unknown environment; One-sided RTT; Wi-Fi and Video Combination.; Databases and Information Systems; Software Engineering
• Online Access: https://ink.library.smu.edu.sg/etd_coll/606; https://ink.library.smu.edu.sg/context/etd_coll/article/1604/viewcontent/GPIS_AY2018_PhD_Hai_Truong_4.pdf
• Institution: Singapore Management University

Indoor localization is important for various pervasive applications, garnering considerable research attention over recent decades. Despite numerous proposed solutions, the practical application of these methods in real-world environments with high applicability remains challenging. One compelling use case for building owners is the ability to track individuals as they navigate through the building, whether for security, customer analytics, space utilization planning, or other management purposes. However, this task becomes exceedingly difficult in environments with hundreds or thousands of people in motion. Conversely, the need to track oneself’s location is also meaningful from the perspective of individuals traversing in crowded spaces. These use cases are pertinent, such as meeting friends or reaching a preferred store in random malls or shopping centers. Nonetheless, addressing these use cases requires solutions that can be applied in unknown environments without pre-existing knowledge of those environments. Consequently, solutions should not necessitate the installation of complex devices, require extensive maintenance efforts, or rely on detailed environmental knowledge. This thesis aspires to achieve localization in practical environments and tackle the challenge of densely populated spaces, such as busy malls, event halls, or crowded supermarkets, where many individuals move across multiple non-overlapping floors or areas. This thesis will address two particularly challenging environments: dense environments with thousands of moving people in non-overlapping areas and unknown environments with no maps, fingerprints, or pre-existing knowledge. Firstly, for dense environments, this thesis proposes a solution that leverages data from various sensors to offer a potential remedy. In particular, I integrate Wi-Fi data with video data and aim to mitigate the individual limitations of each data source. Specifically, location information extracted from video footage compensates for localization errors inherent in using Wi-Fi-based RF signals, wherein precise Wi-Fi MAC IDs are utilized to locate devices across different levels and areas within a building. In general, integrating video and Wi-Fi signals offers significant advantages over using a single sensor, primarily due to the unreliability of using Wi-Fi signals alone, which can exhibit localization errors of 10 meters or more. Additionally, the task of localizing and reidentifying individuals in dense, non-overlapping or disconnected environments becomes exceedingly challenging and computationally demanding when relying solely on video data. The proposed system, named DenseTrack, combines Wi-Fi and video data to enhance Wi-Fi localization and track video objects across disconnected video feeds within such environments. DenseTrack connects devices reported by a Wi-Fi location system with specific video blobs obtained through computationally efficient video data analysis. The experiment results indicate that DenseTrack acquires an average match accuracy of 83% within a 2-person distance, with an average latency of 48 seconds in dense environments. Secondly, in addressing the challenges posed by unknown environments, Wi-Fi fine time measurement (FTM), also known as two-sided round-trip time (RTT), emerges as a protocol offering highly accurate localization within 1-2 meters under conditions where two devices cooperatively measure their signal round-trip time. To achieve that precise localization, the distance calculation utilizes the two-sided RTT, which considers the initial transmission time and the reception of the Acknowledgment (ACK) at the device’s side. It also incorporates the time the transmission signal is received and the subsequent transmission of the ACK from the AP back to the device, referred to as the turnaround time on the AP’s side. This protocol is introduced in IEEE 802.11mc, the third maintenance/revision group for the IEEE 802.11 [62]. However, a notable limitation of two-sided RTT lies in the low deployment rate of Access Points (APs) compliant with the 802.11mc protocol in the market. In response to this limitation, an alternative approach employs a simpler protocol known as one-sided RTT, which focuses on measuring the time difference between message transmissions and acknowledgment receipts solely on the device’s side. Unfortunately, this method does not eliminate the packet processing time on the AP, which can result in significant errors. For example, a processing time as small as 0.1 ms at the AP could induce an error of approximately 30 kilometers in distance measurement due to the speed of light. Consequently, this error must be carefully accounted for in the calculations. This thesis presents empirical findings derived from experiments utilizing one-sided RTT, aiming to assess the feasibility of this approach for indoor localization. It also proposes an initial solution employing a neural network to estimate the distance from smart devices to APs. Notably, this solution exhibits versatility in its applicability across various environments. By addressing the challenges posed by both dense and unknown environments, a comprehensive solution for indoor localization in practical scenarios is achieved.