The Research On Indoor Device-free Passive Human Localization Based On Channel Response

With rapid development of Internet of Things,wireless localization has become one of emerging technologies,which is helpful to location-based intelligent server.Compared with traditional active localization,novel wireless device-free passive human localization can detect,locate and identify entities without any electronic instrument,which can be widely applied into security,equipment protection,people management in factory or home.In indoor environment,wireless signal strength is failed to sense the human presence due to its coarse-grained and influence of indoor multiple-path effect.Now,WLAN technology is rapidly developed,and IEEE 802.11 a/g/n protocol adopts orthogonal frequency division multiplexing,which can offer carrier-level fine-grained channel response information.Channel response includes channel state information of multiple subcarriers that depict the multi-path propagation and is helpful to fine-grained wireless sensing technology,which offers new chance for indoor fine-grained and high-accuracy device-free passive human localization research.Recently,device-free passive localization based on channel response is internationally inchoate.Lots of basic problems exit and wait for solutions,including validity,stability and effectiveness of indoor device-free passive localization based on channel state information on sub-carriers.The paper plans to take advantage of channel state information on sub-carriers to achieve advanced indoor device-free passive localization,which is expected to promote the development of device-free passive localization in our country.The paper focuses on fine-grained single-link device-free passive human localization,mainly conduct the research from the following aspects:Firstly,in order to reduce the cost of site survey and improve the universality of device-free passive localization,the paper constructs a larger number of comparison experiments,and explore to leverage amplitude response information to quantify the sensitivity of link to human movement.Then an advanced and lightweight device-free passive human detection model is put forward and realized,which can reduce the cost of site survey.The results of extensive experiments show the adapted device-free passive human detection can achieve less than 5%false positive and false negative,further prove the validity and effectiveness of amplitude response to human movement.Secondly,some experiments show that amplitude response is poor to sense the human movement with low speed.The paper explores to realize an advanced calibration-free device-free passive human detection based on phase response information.The model can gain the available phase after the random phase from commercial NIC is transferred.The paper develops light and non-calibration device-free passive human detection based on variable coefficient of temporal phase information,which can reduce the system cost and improve the general applicability of device-free passive human detection.Plenty of experiments show that the approach can achieve less than 5%false positive and more than 90%detection accuracy,while prove that the stability of phase response in static environment and validity and effectiveness of phase response in dynamic environment.Thirdly,the amplitude response is firstly adopted to estimate the capability of wireless sensitivity,which is helpful to the decision of receiver locations.The paper further researches device-free passive human localization based on Bayesian classification technique.The range and accuracy of passive localization can be improved,the dead pots can be reduced,when the receiver is placed in high wireless sensing.The experiments made in some testbeds show that when the receiver is located in high-sensing location,the localization accuracy can be improve 21%,and prove that channel response information is valid to device-free passive localization.Finally,based on difference and relation of channel state information on different frequencies,the paper puts forward high accurate indoor device-free passive localization.The paper takes advantage of the frequency selective attenuation of indoor wireless signals,and puts forward two novel single-link-based device-free passive human localization algorithm:Weighted Bayesian classification and Maximum Similarity Metric,which can greatly improve the accuracy of device-free passive localization.Extensive experiments show that the methods can achieve less than lm median localization error.The stability and effectiveness of channel state information for passive localization can be proved by vast experiments.