Research On Access Authentication Technology Based On Zigbee Radio Frequency Fingerprint

With the gradual popularity of Io T devices,various industries have extensively developed various Io T solutions.A large number of Io T devices are vulnerable to attacks when exposed in an uncontrolled environment,and Io T devices with insecure factors are connected to the Io T as a potential threat.As an important physical feature of communication transmitters,radio frequency fingerprints are unique,stable,and difficult to be cloned.They can be used to solve the identity authentication and network access security issues of Io T devices.In this thesis ZigBee equipments are taken as the research object,and the OQPSK physical layer standard in the IEEE 802.15.4 protocol is analyzed,and solutions to the ZigBee signal capture problem under the condition of low signal-to-noise ratio and the influence of frequency offset,symbol timing and other factors on ZigBee RF fingerprint are provided.The unique cyclic shift characteristics of ZigBee signals are analyzed,and a new method of RF fingerprint extraction is proposed.By analyzing the process of the IEEE 802.1X authentication protocol,referring to its architecture and introducing radio frequency fingerprint authentication information,a radio frequency fingerprintbased Io T device access authentication technology is proposed,and the feasibility of this technology is verified through experiments.The main work of this thesis is as follows:1.In order to solve the problem that ZigBee signals are difficult to be captured under the condition of low signal-to-noise ratio,three signal detection algorithms are proposed and their advantages and disadvantages are analyzed.The energy detection algorithm detects by setting a threshold,which is fast but has poor performance under low signal-to-noise ratio conditions;the dual-window ratio detection algorithm judges the start and end points of the effective signal by calculating the ratio of the signal energy in the two observation windows,and the detection effect is excellent,but the appropriate window length needs to be selected;the correlation detection algorithm is based on the characteristics of the signal preamble itself,and the normalized cross-correlation value between the same symbols is greater than the cross-correlation value between the effective signal and the noise.Its performance is similar to the dual-window ratio detection algorithm,but the end point of the effective signal cannot be determined.Monte Carlo experimental results show that the performance of the dual-window ratio detection algorithm is better than the other two algorithms.2.In order to solve the problem of large frequency offset in the synchronization process of ZigBee devices,two coarse frequency offset estimation algorithms are proposed and their performances are compared.The frequency offset search algorithm sets a frequency offset compensation step size and compensates the received signal with a gradually increasing frequency offset compensation value.If the maximum value of the correlation value in the synchronization window exceeds the threshold,it means that the frequency offset value of the received signal has been searched;The spectrum correlation algorithm calculates the offset when the spectrum cross-correlation value of the received signal preamble and ideal preamble reaches the maximum,and multiplies it by the frequency resolution of the spectrum to obtain the coarse frequency offset value.The spectrum correlation algorithm is faster than the frequency offset search algorithm,and the estimated coarse frequency offset value is more accurate.3.In order to solve the problem of the symbol cycle change caused by the unstable crystal oscillator of ZigBee devices,a symbol synchronization tracking algorithm is proposed.According to the decoded symbols,the synchronization point is used as the reference point and the left and right points are the starting point to calculate the correlation value between the synchronized received signal and the ideal symbol within a symbol length.Then the point with the largest correlation value is selected as the accurate starting point of the current symbol,and the starting reference point of the next symbol is determined at the same time.By repeating this process,the purpose of tracking the accurate starting point of each symbol is achieved,thereby achieving symbol synchronization.Experiments on the actual ZigBee signals show that the algorithm can track the exact position of each symbol and has excellent performance.4.The unique cyclic shift characteristics of the ZigBee signal are analyzed,and the crosspower spectral density calculation method is proposed to calculate the square sum of the cross-correlation results of the received symbol and the reference symbol and its conjugate.The method is verified by detailed formula derivation and simulation analysis.The obtained cross power spectral density has the characteristics of data independence and anti-noise.The signal data extracted from 54 ZigBee devices were classified by the cross-power spectral density,and the recognition accuracy reached 97.59%.Theoretical derivation and experimental tests prove that the cross-power spectral density feature selected in this thesis has high stability and high device discrimination,and can be used as radio frequency fingerprint of ZigBee devices.5.With reference to the architecture of the IEEE 802.1X authentication protocol and the concept of port-based access control,a radio frequency fingerprint-based Io T device access authentication system is designed by using radio frequency fingerprint as the identity information of Io T devices.Taking ZigBee equipment as an example,the system is implemented and built,and the security of the system is demonstrated through real attack tests and theoretical analysis.