Research On The Composite Of Wifi Preamble And Radar Chirp Signal

In the WiFi environment,there is a great potential for the application of sensing surrounding environment by extracting the WiFi preamble signal,or recognizing the device-free human bodies(without wearable device)by using their gestures and actions.The topic of this dissertation originates from the items Natural Science Foundation of China "Research on Integration of Communication and Intelligent Sensing".The topic has important theoretical and practical significance.This dissertation mainly investigate the composite technology of wireless WiFi signal and radar chirp sensing signal,and several innovative contributions are completed:1.This dissertation proposes a superposition composite method of WiFi long preamble signal and radar sensing signal.Firstly,WiFi long preamble and short-time chirp signal are superimposed in the time domain at the transmitter by and replacing the original WiFi long preamble.Secondly,on the basis of short preamble signal acquisition at the receiver,target recognition is carried out by using the composite long preamble.Finally,as for the fine channel estimation,the corresponding position chirp is deleted before correlating operation with local long preamble.Thus,the recognition ability of the long preamble can be improved;meanwhile the impact on channel estimation is minimized.2.Because the deletion residue of the time-domain short time superposition composite method has an impact on CE and it has high requirement for sampling rate when recognition,we propose a convolution composite scheme involving a WiFi long preamble and a radar chirp signal.That is,without taking up an extra bandwidth and time slot,the transmitter convolutes the preamble with chirp signal;at receiver to channel estimate,before correlation operation with local long preamble,a deconvolution separation of the receiving composite preamble is made first.Simulation results show that the convolution composite based on receiver deconvolute separation,its channel estimation anti-fading ability is better than that of time-domain superposition method.Then more chirp constituent can be composited,meanwhile the recognition requirement of sampling rate is reduced.The raw echoes for different targets demonstrate it's true that the difference is more obvious when the convolution signal is used.3.To further reduce the influence of separation residue on channel estimation,we propose a frequency domain amplitude composite method.Simulation results show that it has exactly the same accuracy and the same excellent anti-fading performance of channel estimation as original WiFi preamble.The recognition ability of the WiFi preamble signal is improved to a certain extent by using this method,although the method is simplified and the phase information of chirp signal is not integrated.4.When wireless signal recognizes target,it is more common that the target is in a specified area with random position rather than a fixed coordinate point,such as the elderly toilet and bathroom fall detection scene.So we study the different heights human body recognition with random position in a small designated indoor area by using the composite signal as well as original WiFi signal.Finally,the research works of the whole dissertation is summarized.And several valuable research directions of target detection and recognition based on the composite signal are discussed.