Research On LoRa End Nodes Protection System And Method Based On RF Fingerprint

The Internet of Things(Io T)uses technologies including sensors,radio frequency(RF)transceiver,big data and cloud computing to achieve extensive interconnection of things through internet access,rapidly and profoundly affecting all walks of life.However,the potential security problems,especially problems in the security protection of the end node,are predominant in current Io T filed.As a kind of wireless communication technologies,LoRa technology is now widely used in the Io T owing to its advantages of low power consumption and long distance using.However,the access control security problems in LoRa end nodes still need to be solved urgently since the current security strategy based on the link layer identity that LoRa uses by default has been proven to have large security risks and hidden dangers,which seriously hinders the application of LoRa technology on a larger scale.In order to improve the security protection ability of LoRa end nodes,different from traditional cryptographic security technology,this thesis introduces physical layer security technology of RF fingerprint method for identification and access control of LoRa end nodes.Moreover,the influencing factors of LoRa end node fingerprints from both circuit principle and hardware configuration aspects are sorted out,and the feasibility of applying RF fingerprint method to LoRa end nodes has been analyzed and demonstrated in detail.Furthermore,based on the analysis of the standard LoRaWAN network architecture and its security,two LoRa gateway solutions based on RF fingerprints to achieve end node access control and the corresponding system architecture and process methods are proposed.The main work and innovations of this thesis are as follows:1.Since the existing LoRa security strategy based on link layer identification has security loopholes in end node protection,this thesis proposes a LoRa end node protection method based on RF fingerprints.Specifically,the influencing factors of LoRa end node fingerprints from both circuit principle and hardware configuration aspects are sorted out,based on which the feasibility of applying RF fingerprints to LoRa end node access control is verified by extracting frequency offset and cross power spectrum density(CPSD)features to analyze the influence of each factor on RF signals.The LoRa security strategy based on RF fingerprints skips the step of transforming the LoRa end node and it enhances the security of end node protection from the perspective of the physical layer on the premise of practical application to better support the LoRa communication services.2.Aiming at the implementation problem of applying RF fingerprints to LoRa end node access control,this thesis proposes a scheme to improve the design of LoRa gateway based on the analysis of the standard LoRaWAN network architecture and its component modules.Specifically,the RF fingerprints are added to the existing end node access control rules of the gateway,which is made as the end node RF signals are received.All the indicators including the fingerprints are synthesized to identify the LoRa end node,and further processing of the end node can be implemented according to the identification result.This thesis proposes to obtain the LoRa end node signal after preliminary processing from the gateway circuit to make a fingerprint.Compared with setting up a separate module to receive signals to make fingerprints,the proposed method of making fingerprints directly from the end node signal in the gateway circuit can eliminate the problem of asynchronization between the LoRa data packet and the corresponding fingerprint.The scheme proposed above provides a new and more reliable LoRa end node protection scheme under the premise of retaining the communication technical indicators of the gateway and being compatible with the existing network architecture.3.In addition to the scheme to transform the LoRa gateway,this thesis also proposes a scheme to combine universal software radio peripheral(USRP)and microcomputer into a new LoRa gateway in view of the implementation of applying RF fingerprints to LoRa end node access control,specifically,based on the signal acquisition ability of USRP and the ability of microcomputer to extract RF fingerprints and train the classifier to identify.Moreover,this article thoroughly explores the undisclosed LoRa physical layer transmission protocol,and finally achieves the software defined radio(SDR)reproduction of the LoRa physical layer transceiver chain,realizing the relay and forwarding function of the transmission message.The solution to replace the original LoRa gateway takes full advantage of the combination of USRP and microcomputer,improves the performance of RF fingerprint extraction and judgment,and maintains the simplicity of the application of the LoRaWAN network architecture simultaneously.