10 November 2023 in IEEE Internet of Things Journal
The popularity of radio frequency (RF)-based fingerprinting for indoor localization has grown owing to its relatively low cost of equipment deployment and satisfactory accuracy. However, generating a complete radio map by associating unlabeled RF signals with the corresponding location information remains challenging, especially in crowdsourcing-based fingerprinting. In this article, we propose a semi-crowdsourced radio map construction method based on Bluetooth low-energy (BLE) landmarks that harnesses reference points (RPs) in the radio map for coarse localization and facilitates the labeling of location information to WiFi signals. Principally, we acquire RF-received signal strength (RSS) measurements annotating them with location coordinates recorded while a user is walking to provide an efficient method of data collection. Furthermore, we introduce a generative adversarial network (GAN)-based method to increase the amount of training data collected at each RP and reduce human effort by augmenting the fingerprint database. Our proposed method demonstrates promising results, including improved localization accuracy and localization performance comparable to that of traditional site surveys while reducing measurement time and human effort.
21 February 2024 in IEEE Access
Location estimation in indoor environments using radiofrequency (RF) has garnered considerable attention in recent years owing to the widespread adoption of mobile devices. RF-based fingerprinting—a direct approach that allows location estimation based on observed signals—relies on manual surveys during the offline phase to create a radio map with coordinates and RF measurements at multiple locations. The accuracy of RF fingerprint-based localization is related to the number of reference points. However, conventional site survey procedures tend to incur substantial expenses. To alleviate the workload of site surveys and address the challenge of incomplete fingerprint databases, we propose a data-augmentation method to complement existing fingerprint data. Our approach leverages a conditional generative adversarial network with long short-term memory (CGAN-LSTM) prediction model to effectively learn the intricate patterns inherent in the initial training data and generate high-quality synthetic data that align with the underlying data distribution. In an experimental evaluation conducted on a real testbed, our data augmentation framework increased the average localization accuracy by 15.74% compared with fingerprinting without data augmentation. Furthermore, experiments conducted in two typical indoor environments using sparse data highlighted the significant performance enhancement of the proposed approach in reducing localization error and was comparable to state-of-the-art data-augmentation methods.
6 August 2021 in IEEE Access
Smartphone-based indoor navigation systems are becoming increasingly popular in a variety of applications. However, localization accuracy has always been a challenge. The Kalman filter (KF) is a well-known estimation in the Bayesian framework, but can only deal with linear problems and Gaussian models. A particle filter (PF) is another essential estimation tool in a Bayesian system. However, a critical challenge with PF is the problem of particle degradation after resampling. To mitigate the particle degradation problem in PF, unsupervised learning based on k-means clustering is proposed in this paper. It forms clusters of similar particles based on the sum of weights. Also, we present enhancing the PF by utilizing a map constraint and k-means clustering (PFMK) and integrating Bluetooth low energy (BLE) along with pedestrian dead reckoning (PDR) for positioning. BLE and PDR-based positioning with a map constraint lead to an increase in accuracy of at least 20% compared with a traditional PF. Moreover, the proposed unsupervised k-means approach increases the accuracy by an additional 20%, whereas the overall performance of PFMK achieves a mean error of <; 1.5 m in the test environments.
18 October 2023 in Internet of Things
6 March 2023 in Future Generation Computer Systems
1 January 2022 in SSRN Electronic Journal
Owing to the ubiquity of wireless networks, WiFi fingerprinting is widely applied in indoor localization. However, constructing a comprehensive radio map for WiFi indoor localization is labor intensive and time consuming. Additionally, the acquired radio map requires frequent updates owing to the dynamic nature of indoor radio environments. In this study, we aim to contribute to the alleviation of these shortcomings. We proposed a crowdsourcing landmark-based approach for radio map creation and update. First, we utilize Bluetooth low-energy beacons as landmark proximity for labeling the WiFi signals. Second, we employed a deep learning approach to generate a landmark classification model. Third, by using only a few known labeled fingerprints from the detected/classified landmarks, we utilized Gaussian process regression to interpolate a detailed radio map, and thereby, overcoming the drawbacks of sparse training data. Consequently, we can reduce the time consumed and cost associated with creating and updating traditional radio maps. Finally, we evaluated our crowdsource-based indoor localization system via real-time experiments using three different tag devices (smartphones from different vendors). The results demonstrated that the maximum likelihood estimation approach for crowdsourcing landmark-assisted localization can realize a localization accuracy of 5 m at 50% for the three devices. Furthermore, experiments show that the proposed system delivers consistent localization accuracy in a heterogeneous device environment while reducing the offline fingerprinting cost.