Research On Multi-band Equipmentless Target Positioning

Today,location information has become an indispensable basic information in daily lives.And with the advancement of wireless communication technology,positioning technology has also achieved unprecedented development and popularity.GPS positioning technology has become a common means of daily travel,but GPS positioning technology is often not used normally in indoor environments and urban canyons due to the inherent weak penetrability problem of GPS signals.Due to the increasing demand for location information,many wireless location techniques based on civilian communication facilities have emerged in recent years,such as Bluetooth positioning,WIFI positioning and so on.However,these positioning techniques all have a common drawback that need targets to carry corresponding devices to communicate with base stations or anchor nodes.With the development of human society,new demands and new applications about location information emerge continuously.In some specific application scenarios,such as rescue scenes,security monitoring,and elderly care in special circumstances,the target is unable or unwilling to carry any equipment,so the traditional positioning methods will be invalid.To overcome this problem,Device-free Localization(DFL)based on Wireless Sensor Network(WSN)is proposed.DFL technique estimates the target location by monitoring the impact of the target on the wireless link in the area,and does not require targets to carry any positioning devices.Therefore,the DFL technique based on WSN has attracted lots of attentions.The 2.4 GHz band is an Industrial Scientific Medical(ISM)open band and has the advantage that the RSS value changes significantly on target-affected links,therefore most DFL systems are currently established on the 2.4 GHz band.However,there is some signal such as Wifi,Bluetooth,and Zigbee on the 2.4GHz band.The interference among them is serious.Moreover the 2.4GHz band signal has disadvantages such as short transmission distance,poor penetration ability and so on.These factors constrain the positioning performance and use occasion of the DFL system on the 2.4 GHz band.In recent years,in order to overcome the shortcomings of the DFL system on the 2.4 GHz band,some WSN systems in other frequency bands have emerged,such as 800/900 MHz,FM band and so on.However,when using these frequency bands WSN for positioning,the advantage of the 2.4 GHz band signal is lost.Aiming at this problem,this paper makes a deep study about how to use multi-band complementary information based on multi-band DFL system to entance DFL positioning performance.The main work and research results of this thesis include the following aspects:(1)The multi-band DFL system is independently designed and developed.The system consists of the 2.4 GHz band,the 433 MHz band and the 915 MHz band WSN.The 2.4 GHz WSN is constructed with the ZigBee module CC2530.WSNs in 433MHz and 915MHz bands are based on STC90LE58RD+MCU and Si4463 module.The three sets of WSNs all adopt the cyclic mechanism to realize the networking function.(2)This thesis designs a multi-band Radio Tomography Imaging(RTI)fusion and target detection method.Due to the different environmental noise interferences in each frequency band,the RTI results presented in each frequency band are also different.From this point of view,this thesis designs a multi-band RTI image weighting fusion method based on communication quality information,which reduces the interference of pseudo-targets and background noises to some extent.In order to deal with the influence of the unavoidable pseudo-targets and background noises on the RTI results,this thesis also designs a RTI target detection method based on Convolutional Neural Network(CNN),which effectively improves the positioning accuracy.(3)This thesis proposes a multi-band DFL correlation matching pursuit algorithm based on joint sparse model.By analyzing the high correlation among the atoms in the DFL weight dictionary under Compressive Sensing(CS)framework and the limitations of traditional solutions,a correlation matching pursuit algorithm which can utilize the inter-atomic correlation between dictionaries is designed.On the basis of this algorithm,and inspired by Joint Sparse Model(JSM),a multi-band DFL fusion localization algorithm based on the compressed sensing framework is proposed.Finally,the research work of this thesis is summarized,and the shortcomings in the existing work are discussed.The improvement suggestions are put forward in order to improve in the future research.