Due to the emergence of various wireless sensing technologies, numerous positioning algorithms have been introduced in the literature, categorized intogeometry-driven positioning(GP) anddata-driven positioning(DP). These approaches have respective limitations, e.g., a non-line-of-sight issue for GP and the lack of a high-dimensional and labeled dataset for DP, which could be complemented by integrating both methods. To this end, this paper aims to introduce a novel principle calledcombinatorial data augmentation(CDA), a catalyst for the two approaches’ seamless integration. Specifically, GP-based data samples augmented from different positioning element combinations are calledpreliminary estimated locations(PELs), which can be used as high-dimensional inputs for DP. We confirm the CDA’s effectiveness from field experiments based on WiFiround-trip times(RTTs) andinertial measurement units(IMUs) by designing several CDA-based positioning algorithms. First, we show that CDA offers various metrics quantifying each PEL’s reliability, thereby extracting important PELs for WiFi RTT positioning. Second, CDA helps compute the observation error covariance matrix of a Kalman filter for fusing two position estimates derived by WiFi RTTs and IMUs. Third, we use the important PELs and the above position estimate as the corresponding input feature and the real-time label for fingerprint-based positioning as a representative DP algorithm. It provides accurate and reliable positioning results, with an average positioning error of 1.58 (m) and a standard deviation of 0.90 (m).

This paper introduces a novel approach, mobility-induced graph learning (MINGLE), to enhance the accuracy of Wi-Fi positioning. Traditional Wi-Fi positioning methods often struggle with accuracy due to obstructions and interference. MINGLE addresses these challenges by converting user movement patterns into graphs, which are then analyzed using graph neural network. This method involves creating two types of graphs, based on the time and direction of user mobility, and employs a novel cross-graph learning technique in conjunction with self-supervised learning. This approach has demonstrated significant improvements in positioning accuracy, achieving a remarkable accuracy of 1.301 (m) in an underground parking lot setting, without relying on labeled data samples.

A smartphone-based user mobility tracking could be effective in finding his/her location, while the unpredictable error therein due to low specification of built-in inertial measurement units (IMUs) rejects its standalone usage but demands the integration to another positioning technique like WiFi positioning. This paper aims to propose a novel integration technique using a graph neural network called Mobility-INduced Graph LEarning (MINGLE), which is designed based on two types of graphs made by capturing different user mobility features. Specifically, considering sequential measurement points (MPs) as nodes, a user’s regular mobility pattern allows us to connect neighbor MPs as edges, called time-driven mobility graph (TMG). Second, a user’s relatively straight transition at a constant pace when moving from one position to another can be captured by connecting the nodes on each path, called a direction-driven mobility graph (DMG). Then, we can design graph convolution network (GCN)-based cross-graph learning, where two different GCN models for TMG and DMG are jointly trained by feeding different input features created by WiFi RTTs yet sharing their weights. Besides, the loss function includes a mobility regularization term such that the differences between adjacent location estimates should be less variant due to the user’s stable moving pace. Noting that the regularization term does not require ground-truth location, MINGLE can be designed under semi- and self-supervised learning frameworks. The proposed MINGLE’s effectiveness is extensively verified through field experiments, showing a better positioning accuracy than benchmarks, say mean absolute errors (MAEs) being 1.510 (m) and 1.077 (m) for self- and semi-supervised learning cases, respectively.

Recently, round-trip time (RTT) measured by a fine-timing measurement protocol has received great attention in the area of WiFi positioning. It provides an acceptable ranging accuracy in favorable environments when a line-of-sight (LOS) path exists. Otherwise, a signal is detoured along with non-LOS (NLOS) paths, making the resultant ranging results different from the ground truth, called an RTT bias, which is the main reason for poor positioning performance. To address it, we aim at leveraging the user mobility trajectory detected by a smartphone’s inertial measurement units, called pedestrian dead reckoning (PDR). Specifically, PDR provides the geographic relation among adjacent locations, guiding the resultant positioning estimates’ sequence not to deviate from the user trajectory. To this end, we describe their relations as multiple geometric equations, enabling us to render a novel positioning algorithm with acceptable accuracy. Depending on the mobility pattern being linear or arbitrary, we develop different algorithms divided into two phases. First, we can jointly estimate an RTT bias of each access point (AP) and the user’s step length by leveraging the geometric relation mentioned above. It enables us to construct a user’s relative trajectory defined on the concerned AP’s local coordinate system. Second, we align every AP’s relative trajectory into a single one, calledtrajectory alignment, equivalent to transformation to the global coordinate system. As a result, we can estimate the sequence of the user’s absolute locations from the aligned trajectory. Various field experiments extensively verify the proposed algorithm’s effectiveness that the average positioning error is approximately 0.369 (m) and 1.705 (m) in LOS and NLOS environments, respectively.

As the number of smartphone users exploded, the demand for Location-Based Service (LBS) has increased. It is important for the LBS to specify the user location by utilizing the sensor built in the smartphone. Unlike outdoor localization, which can employ GPS, there are many challenging issues in indoor localization including non-line-of-sight (NLOS) and multipath effect. In our paper, we focus on Wi-Fi Fine Timing Measurement (FTM) which is a new function of the Android Pie Operating System (OS). We propose line-of-sight (LOS) identification algorithms applicable to Wi-Fi FTM and apply these algorithms to indoor localization based on multilateration methods. We utilize a hypothesis test framework and Support Vector Machine (SVM) to identify LOS signals. We divide LOS/NLOS signals as low and high-quality signals according to the degree of multipath error. We achieve high-quality signals identification rate of 92.4% on average in the sample size 99 and of 78.3% on average in the sample size 29. Therefore, we obtain a 24.4% localization performance improvement compared to the perfect LOS detector by using only high-quality signals to localization.