Experimental Investigation of Body-Centric Indoor Localization Using Compact Wearable Antennas and Machine Learning Algorithm
Abstract— A simple and effective body-centric localization algorithm has been proposed in this article using ultra-wideband (UWB) wearable technology. The algorithm has been validated through measurement campaigns with a human subject volunteer in an indoor environment. The algorithm takes into account statistical channel parameter analysis, machine learning (ML) algorithms, and time-of-arrival (TOA)-based range estimation/data fusion techniques. Two channel parameters namely path loss magnitude and rms delay spread are proposed as classification features to be applied to the ML algorithm to accurately classify the off-body channel links into LOS, PNLOS, and NLOS scenarios. Multiclass support vector machine (MC-SVM) classifier along with SMOTE algorithm to take into account class imbalance is applied with the channel classification accuracy of 98.63%. Threshold-based range estimation algorithms are applied in order to mitigate NLOS scenarios caused mainly due to the presence of the human subject. Results report human localization accuracy in 0.5–3 cm range using TDOA data fusion technique for target estimation. Further validation is presented considering wide range of Tx–Rx distance, presence of another obstruction between the Tx and Rx links, and performance in different environment which shows the suitability of the proposed methodology. Index Terms— Body-centric communication, channel characterization, indoor localization, machine learning (ML), wearable antennas.
I. INTRODUCTION
BODY-CENTRIC indoor tracking and localization enables a plethora of applications in the day-to-day, healthcare, military, industry, and entertainment sphere. With the advancement in technology, miniaturization of wearable antennas/sensors and the application of machine learning (ML) algorithms have gained tremendous importance to provide Manuscript received January 7, 2021; revised July 8, 2021; accepted July 20, 2021. Date of publication September 15, 2021; date of current version February 3, 2022. (Corresponding author: Richa Bharadwaj.) This work involved human subjects or animals in its research. The authors confirm that all human/animal subject research procedures and protocols are exempt from review board approval. Richa Bharadwaj and Shiban K. Koul are with the Microwave and RF Group, Centre for Applied Research in Electronics, IIT Delhi, New Delhi 110016, India (e-mail: richab@care.iitd.ac.in; shiban_koul@ hotmail.com). Akram Alomainy is with the School of Electronic Engineering and Computer Science, Queen Mary University of London, London E1 4NS, U.K. (e-mail: a.alomainy@qmul.ac.uk). Color versions of one or more figures in this article are available at https://doi.org/10.1109/TAP.2021.3111308. Digital Object Identifier 10.1109/TAP.2021.311130
bust communication links and information for various wireless body area network (WBAN) applications at a large scale. Requirement of multifunctional WBAN systems with low-cost, low-power have led to the development of various communication technologies such as Bluetooth, Wi-Fi, UWB, and ZigBee [1]–[3]. In recent years, ultra-wideband (UWB) communication has gained popularity for body-centric wireless communication applications due to its large bandwidth (sub-ns pulses) [4], [5]. It offers attractive features such as compact size, high data rate, low power, low cost, robustness against multipath, penetration through obstacles, integration with other technologies [6], [7]. This makes UWB an ideal candidate for compact body-centric wireless communication systems.
The development of robust wireless localization algorithms and techniques for location based services has attracted significant interest in the academic, industrial, and research domains [8]. Position estimation algorithms generally use methods such as time of arrival (TOA), time difference of arrival (TDOA), angle of arrival (AOA), and received signal strength (RSS) and also hybrid measurements for accuracy improvement [9]. In order to save space and cost of the localization system setup, limited number of base stations (BSs) are generally deployed which act as the anchor nodes
Apart from direct path propagation between the transmitter and receiver which is represented as line-of-sight (LOS) condition, non-LOS (NLOS), partial/pseudo-P(NLOS) scenarios occur in a cluttered indoor environment which deteriorates the performance of the UWB positioning system. NLOS scenarios occur due to obstruction caused by objects, furniture, walls, glass, and also human subjects in the room [10]–[13], leading to signal attenuation, high multipath and positive bias error in the range estimates. PNLOS scenario refers to partial-NLOS in which the direct path between the Tx and Rx is partly obstructed leading to high multipath channel [6], [14], [15] or pseudo-NLOS in which visually there seems to be total obstruction, but the first path is distinctly detectable with high number of multipath components (MPCs) [16]. The human body is a complex environment for the propagation of a wireless signal, and serves as one of the major obstacle for propagation of the signals [12], [13], [17]. Hence, it is important to understand and characterize the body-centric radio channel for enhancing the performance of the wearable.
0 Comments