Research On Gesture Recognition Method Based On WiFi Signals In The Environment With Interference

At present,WiFi has become more and more popular in our lives and is gradually applied to the Internet of Things perception.By measuring the influence of human activities on the amplitude and phase of WiFi wireless signal,gesture recognition and gait pattern recognition can be completed.However,most of the current research on gesture recognition based on WiFi signal mainly considers the ideal scene,that is,only the target is active in the recognition scene,and there is no other dynamic interference.This limits the utility of activity recognition based on WiFi signals.Especially in smart home or smart office environments,in addition to the identified targets,there are usually other people present.Therefore,this thesis focuses on the method of gesture recognition in the environment with interference and proposes a corresponding schemes.Specifically,this thesis first determines whether there is interference behavior from other people in the surrounding environment.If there is interference,the classification algorithm based on cross-correlation is used to eliminate the influence of other people's activities in the surrounding environment and classify the target behavior.The commercial WiFi devices are utilized to evaluate the performance of the system and demonstrate the effectiveness of the preposed scheme in the typical indoor environment.In addition,this method does not require specialized hardware to eliminate the interference introduced by the surrounding people and can be implemented on commercial WiFi devices.