A dual-polarized Doppler Radar system for fall detection in an indoor environment

点击翻译按钮获取中文摘要

 DESCRIPTION (provided by applicant): Falls are a leading cause of injuries and deaths among older adults (OAs), and will be an increasing public health concern given projections of future increases in the number of OAs in the US. Fall detection and/or surveillance systems are needed to enhance an OA's confidence and independence, and to facilitate rapid post-fall responses. This project will develop a low-cost fall detection system, to facilitate passive and wireless fall detection in diverse indoor environments, and which can be easily translated into a market product. Central to this system is the use of dual-polarized Doppler radar, an enhancement over existing radar systems for fall detection, and which also overcomes several limitations of alternative approaches (e.g., dependency on patient compliance with wearable sensors, privacy concerns with video-based systems). This project involves three Aims. In Aim #1, we will develop the system by designing a dual-polarized antenna for application to realistic indoor use, using a circularly-polarized, stub-loaded helix antenna for transmitting and a dual-linearly-polarized receiving antenna. Performance of the designed antenna system will be assessed, then integrated into a commercial software-defined radio, and subsequently tested to characterize signal detection range and angular coverage and to determine the best installation location. In Aim #2, experimental data will be obtained for simulated falls and activities of daily living, from among a large and diverse sample of human participants. Using these data, in Aim #3 we will develop and evaluate a new fall detection algorithm. This algorithm will employ multiple classification technologies and decision fusion methods to benefit from the rich information provided by the proposed radar system. The current project is novel in the use of dual-polarization, and is targeted to harness the upcoming spectrum- sharing band. It is innovative in taking advantage of polarization diversity in a fall detection algorithm, which is expected to decrease false alarms, and also in emphasizing the discrimination of critical vs. non-critical falls. An extensive experimental database will be developed, including a wider range of activities and events (including slip/trip induced falls), and broader representation of inter-individual diversity, than in many prior studies. By design, the proposed system is expected be an effective, low-cost pervasive indoor fall detection system, with a small form factor, and that provides less intrusive, richer, and more reliable information (e.g. low false alarm rate) than alternative approaches for fall detection.