Omnidirectional gait detection algorithm based on channel state information (CSI)

Abstract

The invention discloses an omnidirectional gait detection algorithm based on channel state information (CSI). The omnidirectional gait detection algorithm comprises the following steps that firstly, a relation model MOFS between movement speeds of different parts of a human body and Doppler frequency shift is created, and WiFi signal frequency change components caused by footstep movement are extracted from mixed CSI by using the model; then, frequency spectrum information is analyzed on a time axis by using a frequency spectrum segmentation technology based on frequency energy slicing, and the instantaneous radial step speed of each measurement point is obtained by using an energy cumulative percentile method according to the MOFS model; and finally, a relation model PLCR-Dir of a WiFi signal propagation path length change rate and the walking direction is put forward, and the calculation of the walking period, the step speed and the step length in all directions is realized through the model. Experimental results show that compared with an existing gait detection algorithm, the gait detection algorithm based on the CSI is smaller in detection error, and three gait parameters including the walking period, the step speed and the step length can be detected at the same time.

Description

Technical field[0001]The present invention relates to the field of wireless networks, and more particularly to dynamic perceived techniques in a wireless network, specifically a full-direction genometric detection algorithm based on CSI.Background technique[0002]In human behavior activities, walking is a complicated behavior controlled by the human brain, and the gait is the posture and action presenting when people walk, can use the three parameters of the step size, step rate, and walking cycle. Assess a human gait. In terms of gait detection, the more detection method is based on the sensor's gait detection and visual analysis. The WiFi signal is used to detect only the normal home router, and the perceived goals do not need to carry any devices, and only the subjects can be active within the WiFi coverage, this method has better applicability. Based on human gait detection applications include indoor positioning and navigation, human identification, patient recovery training, he...

AI Technical Summary

Problems solved by technology

Although gait detection based on sound waves and radar is not affected by light, clothing and other conditions, both require special ultrasonic or radar wave transmitters and receivers, which limits the application scenarios of this technology

[0010] Based on CSI gait detection technology, the calculation of gait information is relatively simple, such as the calculation of a single step length or a separate walking cycle, etc.

A single gait parameter will have a greater impact on the accuracy of gait-based identification

Also in medical treatment, a single gait parameter data cannot predict the patient's physical condition very well, and multiple gait parameters need to be used for joint analysis

At present, there is no quantitative model that correlates the speed change of different parts of the body with the Doppler frequency shift, and the physical law between the received CSI signal change and the gait cannot be represented through a mathematical relationship

Images