Computational Framework for Human Detection Through Improved Ultra-Wide Band Radar System

Abstract

This paper presents a framework for human detection using an ultra-wideband (UWB) radar system and proposes a novel UWB radar antenna design with double-winding structures for radar applications. The proposed antenna achieves high gain and bandwidth, overcoming the shortcomings of Vivaldi antennas, which are the preferred antennas for radar applications. In the proposed novel design, winding structures are incorporated nearer to the main resonator, which suppresses harmonic distortion and enhances the radar detection process. The proposed antenna is designed to cover a frequency range from 6 GHz to 10.5 GHz. The design is analyzed for different design configurations with and without winding structures. However, the dual-side winding structure configuration achieves better performance than the single-side winding structure. The proposed design achieves a maximum bandwidth of 2.09 GHz in the dual-side winding structure configuration, which is a remarkable achievement for radar and precise locating applications. The proposed design achieves improved return loss performance greater than -10 dB in all operating frequencies and voltage standing wave ratios equal to 1 with improved gain and directivity.

An improved positioning method based on compensation and optimization of ultra-wideband ranging results

“…Scholars [19] have introduced an Extended Kalman Filter-based attitude computation method, alongside a Zero-Update of Velocity-based successive gait segmentation method, tailored to mitigate the errors prevalent in indoor pedestrian navigation, thereby improving both the accuracy and robustness of positioning. Additionally, other researchers [20] have proposed a human body detection framework utilizing UWB radar systems, complemented by a uniquely designed UWB radar antenna featuring a dual-winding structure, significantly refining localization precision. In the realm of Kalman filter research [21], there are scholars employing deep learning techniques to finetune the necessary parameters, adapting the filter for various scenarios.…”

Attitude solving aided UWB/IMU integrated algorithm with real-time NLOS suppression

“…Its applications span across a broad range of domains, including indoor navigation, intelligent medical technology, and smart homes [2, * Author to whom any correspondence should be addressed. 3]. However, the intricate indoor environments, coupled with the complex multipath effects that obstruct UWB channels, give rise to non-line-of-sight (NLOS) errors, thereby rendering independent UWB indoor positioning a formidable challenge [4][5][6].…”