Security Evaluation And Enhancement Of Wireless Channel Secret Key

With the commercialization of 5G mobile communication and the research on 6G mobile communication,the rate and coverage of wireless communication will be further improved,which brings higher security requirements for wireless communication systems.The traditional security method is to ensure the security of the system by computing security.However,with the advent of the Internet of Everything era,the distribution of the secret key will bring huge overhead.Moreover,due to the development of quantum computer,the security of the traditional cryptography based on computing complexity will be breached more quickly.Wireless channel secret key makes use of the characteristics of wireless channel to generate the secret key,can avoid the distribution of the secret key through the third party,and provides security based on information theory,so it has become a hot research topic in recent years.In the case of rich scatters and movements,the key generation method of wireless channel has been widely studied.However,the key generation performance of existing methods is not ideal in weak security scenarios such as nearby eavesdropping and quasi-static channel.In addition,the existing methods mainly focus on passive attack,neglecting to defend against active attack.Based on the above problems,the main work and innovations of this paper are as follows:(1)The variation of channel correlation with distance is evaluated and a key generation method based on amplitude remainder is proposed.Aiming at the problem that multi-scene channels are vulnerable to nearby eavesdropping due to lack of decorrelation,this paper analyzes the relationship between the correlation coefficients of eavesdropping channels and legitimate channels with distance variation under four typical indoor and outdoor scatter models,indicating that the theory of half-wavelength decorrelation in many scenes is not valid.In this paper,a key generation method of wireless channel based on channel amplitude remainder is proposed,which makes the characters of legal channel tend to be evenly distributed and widens the difference between them and those of eavesdropping channel.Simulation and experimental results show that the method can pass the randomness test,and the BER of eavesdropper is increased to about 0.5,while the BER of legitimate user is kept below 0.2.(2)The method of increasing quasi-static channel variation is evaluated,and the key generation method based on channel obfuscation is proposed.In the quasi-static channel,channel characters change slowly,and the existing methods have the problems of low KGR and randomness.In this paper,the problems and challenges of antenna cycle and random scheduling are investigated experimentally,as well as the potential attacks in quasi-static channels.A wireless channel key generation method based on channel obfuscation is proposed.This method obfuscates the channels by random filter and antenna scheduling,allows the legal parties to obtain the same channel characters,and prevents the third party from inferring the obfuscation information,and extracts the secret key through K-L transform and adaptive quantization.Experimental results show that the KGR can reach 67bit/pkt,and the generated secret key can pass NIST randomness test.In addition,the proposed method can effectively defend against four potential attacks in static quasi-channels.(3)The threat mode of active attack is evaluated and a the key generation method based on uncooperative frequency hopping is proposed.Aiming at the problem that the existing key generation method is difficult to defend against the active attack,this paper models the ideal attack and the non-ideal attack,analyzes the realization probability of the two attacks and the threat to the system.A the key generation method based on uncooperative frequency hopping is proposed.By analyzing the arrival signal in a time slot,the method determines whether there is an attacker and the time of frequency hopping,and improves the hit ratio of random frequency hopping by setting the different frequency hopping rate.Simulation results show that under active attack,the method can significantly reduce the BER and reduce the attacker KRR to about 0.In addition,the number of key generation in 1000 time slots can reach more than 250 bits.