Research On ZigBee Device Recognition Based On Software Defined Radio

With the development of science and technology,people’s living standards continue to improve,and the demand for home intelligence is also increasing.Io T devices are an indispensable part of home intelligence.The extensive use of Io T devices constitutes a huge impact on users’ privacy and network security.threaten.Among them,ZigBee protocol is widely used in smart home,criminals use ZigBee sensors to collect the living habits of victims and then commit crimes.In order to reduce the security risk of the home,it is necessary to sniff and analyze the surrounding malicious ZigBee devices,so as to protect the privacy and property security of users.However,there are still many deficiencies in the sniffing analysis of the ZigBee protocol,which are embodied in: 1.The sniffing analysis of the ZigBee protocol is mostly based on a proprietary chip platform,which cannot obtain the characteristics of radio signals.2.Most of the current identification methods for ZigBee devices are based on the characteristics of protocol packets,and there is a lack of identification methods that integrate the two characteristics.3.There are few ZigBee protocol sniffing systems in the existing market,and the sniffing analysis of ZigBee protocol is very necessary for researchers.The main work of this thesis is as follows:1.Propose a ZigBee sniffing and identification method based on software radio.First,the ZigBee protocol is sniffed and analyzed by software radio,and the characteristics of ZigBee packets are extracted for device identification.Then,a hybrid classification method is proposed for the situation that the recognition effect is unstable.The classification method uses two kinds of features to classify,one is the parameters in the network training process,and the other is the feature distance between the test sample features and the data set sample features..Finally,the effects of different classification algorithms are compared and analyzed.2.Propose a ZigBee RF fingerprint identification method.Firstly,the mechanism of RF fingerprint generation is studied and compared,and the principles and effects of different fingerprint extraction methods are compared.Secondly,the comparative analysis of radio frequency fingerprint identification methods.Finally,it is proposed to fuse data packet features and RF fingerprint features to improve the classification effect.3.Design and implement the ZigBee protocol sniffing system based on software radio.The system involves both hardware and software systems,in which the hardware system takes LimeSDR as the core,and the software system adopts B/S structure.It can realize the sniffing and identification of common protocols including Wi-Fi,Bluetooth,ZigBee and other protocols,and has the characteristics of expanding and developing other protocols.