Font Size: a A A

Study On Secret Key Generation Schemes Of Physical Layer

Posted on:2022-07-09Degree:MasterType:Thesis
Country:ChinaCandidate:J Y ZhaoFull Text:PDF
GTID:2518306605467454Subject:Master of Engineering
Abstract/Summary:PDF Full Text Request
The physical layer key generation is an information security technology that uses wireless channel characteristics as random source to generate secret key.The characteristics of wireless channel mainly include reciprocity,spatial uniqueness,time variation and unpredictability.These inherent characteristics complement each other to ensure the safe conduct of wireless communication.The physical layer key generation technology does not require a trusted third party for key distribution and management.It not only overcomes the unbreakable assumption of many difficult mathematical problems in traditional encryption schemes,but also greatly improves key generation rate.Based on the existing physical layer key generation schemes,this paper proposes two new physical layer key generation schemes for possible problems in these existing schemes.(1)In the single-carrier system,a highly consistent key generation scheme based on the second modified level crossing algorithm(SMLCA)is proposed,which mainly solves the problem of high key agreement rate(KAR)but low key generation rate(KGR)in existing key generation schemes.First,in order to smooth noise,the weighted sliding window method is adopted to weight measured value.Secondly,this scheme uses K-means clustering to quantify and classify the measured values.K-means clustering algorithm can make full use of both amplitude and phase information.Thirdly,two improved algorithms based on the basic level crossing algorithm(BLCA)are proposed,including first modified level crossing algorithm(FMLCA)and second modified level crossing algorithm(SMLCA).Simulation shows that three negotiation algorithms can effectively reduce bit error rate(BER),and SMLCA has the lowest BER and the highest KGR.Therefore,this scheme uses SMLCA to negotiate bit stream.Furthermore,cyclic interleaving quantization is introduced to solve the problem of the deterioration of key randomness(KR)caused by negotiation algorithm.By increasing the interleaving dimension,KGR can be linearly increase.Finally,by comparing with existing key generation schemes,the superiority of proposed scheme is verified.(2)Under the OFDM system,a highly random key generation scheme based on multi-bit adaptive phase quantization is proposed.First,in the preprocessing stage,we perform Karhunen-Loève Transformation(KLT)to remove the data correlation between sample vectors.Second,in the quantization stage,we propose multi-bit adaptive phase quantization(MAPQ),which can dynamically determine the number of quantized bits based on the correlation between the channel measurement values.Then,lossy multi-bit adaptive phase quantization(LMAPQ)algorithm based on the MAPQ is proposed.Next,we compare adaptive amplitude quantization and adaptive phase quantization.The result show that compared with adaptive amplitude quantization,adaptive phase quantization has lower computational complexity.LMAPQ algorithm discard so much sample information that KGR is decreased.Therefore,through comprehensive comparison,this scheme adopts MAPQ algorithm in the quantization stage.Finally,randomness extractor is proposed in order to obtain perfect secret key,and NIST test is used to verify the effectiveness of randomness extractor.
Keywords/Search Tags:Key generation, Weighted sliding window, Negotiation algorithm, Cyclic interleaving quantization, KLT, Multi-bit adaptive quantization, Randomness extractor
PDF Full Text Request
Related items