Indoor Localization Based on Wi-Fi Received Signal Strength Indicators: Feature Extraction, Mobile Fingerprinting, and Trajectory Learning

Abstract

This paper studies the indoor localization based on Wi-Fi received signal strength indicator (RSSI). In addition to position estimation, this study examines the expansion of applications using Wi-Fi RSSI data sets in three areas: (i) feature extraction, (ii) mobile fingerprinting, and (iii) mapless localization. First, the features of Wi-Fi RSSI observations are extracted with respect to different floor levels and designated landmarks. Second, the mobile fingerprinting method is proposed to allow a trainer to collect training data efficiently, which is faster and more efficient than the conventional static fingerprinting method. Third, in the case of the unknown-map situation, the trajectory learning method is suggested to learn map information using crowdsourced data. All of these parts are interconnected from the feature extraction and mobile fingerprinting to the map learning and the estimation. Based on the experimental results, we observed (i) clearly classified data points by the feature extraction method as regards the floors and landmarks, (ii) efficient mobile fingerprinting compared to conventional static fingerprinting, and (iii) improvement of the positioning accuracy owing to the trajectory learning.Wi-Fi RSSI based indoor localization [1-4] is one of the standard approaches for indoor localization. It is able to utilize the RSSI measurements received from a large number of access points (APs) that are already built in construction. However, the Wi-Fi RSSI as a function of distance between a receiver (smartphone) and a transmitter (wireless AP) is nonlinear and varying due to interference of the indoor environments such as the other radio signals, walls, and obstacles. To address this problem, many machine learning based localization methods [5-8] have been developed, which learn a pattern of the RSSI measurements corresponding locations across the interested positioning area. In addition, due to its unbiased estimation capability, it is likely to be combined with other kinds of localization, such as pedestrian localization using inertial measurement unit (IMU) [9,10], visual localization [11,12], and magnetic sensor-based localization [13,14].In particular, semisupervised learning algorithms have been recently suggested for efficient indoor localization, which reduce the human effort necessary for collecting training data [15][16][17][18][19][20]. For example, for indoor localization, a large amount of unlabeled data can be easily collected by recording only Wi-Fi RSSI measurements without requiring position labels, which can save resources for collection and calibration. By contrast, labeled training data have to be created manually. Adding a large amount of Appl. Sci. 2019, 9, 3930 2 of 21 unlabeled data in the semisupervised learning framework can prevent the decrement in the estimation accuracy that occurs when using only a small amount of labeled data.Given the advantage of semisupervised learning, this study describes (i) feature extraction, (ii) mobile fingerprinting, and (iii) mapless loc...

WLAN RSS-Based Fingerprinting for Indoor Localization: A Machine Learning Inspired Bag-of-Features Approach

“…CWT transformation is an additional stage before feature extraction that improve precision at the cost of the computational complexity of the model. Other works can be found in the literature using principal component analysis for feature extraction [ 31 , 32 ].…”

DNN-Based Indoor Localization Under Limited Dataset Using GANs and Semi-Supervised Learning

“…The resulting pseudo-labeled samples are used subsequently to enlarge the labeled data set, i.e., providing additional training information, to build a more general model. Pseudo-labels can be determined based on solving optimization problems as described in [24]- [27] or applying a supervised DL model on the labeled data as described in [28]. However, pseudolabeled data can penalize the performance if they are not introduced to the training in an appropriate way.…”

Automated Data Acquisition in Construction with Remote Sensing Technologies

“…It is done by tracking the changes in the received electromagnetic wave patterns and compare them with calibrated models to count the number of workers or vehicles on site, and to provide rough information about their location and behavioral algorithms. However, due to some legal restrictions for using smart phones on site and not fully accurate results of this technology its use in construction industry is not common yet [17][18][19]. In a different context, barcode and radio frequency identification devices (RFID) proven to be useful technologies for material identification with applications in tracking resources for a supply chain management system [20,21].…”