CSI METHOD FOR RECOGNIZING HUMAN FALL IN WI-FI INTERFERENCE ENVIRONMENT

摘要

The present invention relates to a method for identifying a human fall by dynamic sub-carrier selection of a Wi-Fi CSI in a Wi-Fi interference environment, belonging to the technical field of wireless communications. In the method, CSI interference intensity and a CSI active ratio are first analyzed, a Wi-Fi interference feature mapping matrix is constructed, and the matrix is used to calculate an interference index of each channel to achieve interference discrimination. Then, by means of a dynamic subcarrier selection algorithm CSI-DSSA based on the interference index, the subcarrier combination having the weakest cross-correlation in the interference data is selected for interference processing, and a multi-link data fusion CSI-MLDF method is analyzed to aggregate the time-domain characteristic information of the multiple data streams in undisturbed data. Finally, the time-domain feature value is extracted and a SVM multi-activity classification model is constructed in a Wi-Fi interference environment to obtain a fall action recognition result. The present invention can effectively improve the recognition accuracy of human falling action in a Wi-Fi interference environment.