Research On Dynamic Covert Communication Strategies For Full-duplex Cognitive Networks

Communication security has long attracted wide attention.Especially in the case of large-scale deployment of Internet of Things(Io T)devices,Io T nodes that penetrate into all as-pects of life are more likely to become objects of eavesdropping and data interception.As a key technology to guarantee the security of civil and military communication,covert com-munication plays an important role in various communication fields due to its high secu-rity and reliable transmission characteristics.Existing covert communication technologies provide strong security guarantee for communication behavior by taking advantage of the uncertainty of transmitted signals and the uncertainty of the electromagnetic environment at the eavesdropper's receiver.However,on the one hand,Io T nodes have energy constraints and the complexity of node deployment.While existing covert communication technologies need to sacrifice communication rate for covert performance,which makes it difficult to meet Io T devices'covert performance and transmission rate requirements at the same time.At home and abroad,on the other hand,new theory of the signal detection,analysis,and de-cipher emerge in endlessly.For example,attacks on existing mature communication system has become sophisticated,but due to fixed single strategy and maladjustment of the electro-magnetic environment,the available covert communication technologies are unable to meet the increasing communication security requirements.To satisfy the communication security of Io T system and improve the capacity of covert communication,it is urgent to explore new theories and strategies of covert communication based on the cognition of electromagnetic environment.This thesis mainly focuses on a cognitive covert communication strategy,from the detec-tion scheme and covert scheme of cognitive covert communication to carry out the specific research.Firstly,a real-time detection scheme of“Spectrum Shelter”based on High-order Cumulants(HOC)is proposed,which can obtain the real-time occupancy of authorized fre-quency band or public frequency band,and select the appropriate“Spectrum Shelter”based on it.This detection scheme is robust to the uncertain noise.Then,a covert communication scheme based on Improper Gaussian Signaling(IGS)is proposed,and the covert strategy is dynamically adjusted according to the characteristics of warden and“Spectrum Shelter”.By optimizing the transmitting power and circular coefficient of IGS,the covert transmission rate can be improved while the covert performance is protected.Finally,the performance of the proposed scheme is verified by sufficient simulations.The main research work and innovative achievements of this paper are summarized as follows:1.The HOC based real-time detection scheme of“Spectrum Shelter”is studied.Ac-cording to the number of sensing antennas of full-duplex cognitive covert user,two sensing schemes are proposed.When the cognitive covert user is equipped with a single sensing antenna,the rich statistical features of the received signal samples are firstly extracted through HOC.Then the hypothesis testing model is divided into two cases according to whether the cognitive covert user is in communication state.In each case,the Likelihood Ratio Test is used to construct the test statistics,and the ex-pressions of the detection probability and false alarm probability are given according to the distribution of the test statistics.Besides,the closed solution of the decision threshold is derived by Neyman-Pearson(NP)criterion.When the cognitive covert user is equipped with multiple sensing antennas,the sample's HOC is also obtained first.Then the hypothesis testing model is divided into two cases according to the communication state of the cognitive covert user.In each case,the Hotelling T~2s-tatistic is used to construct the test statistic according to the statistical characteristics of the received signal's HOC.After that,the closed solution of the decision threshold is derived by NP criterion,and the estimated detection probability is obtained ac-cording to the Gaussian asymptotics of the non-central F-distribution.The theoretical computational complexity of the two proposed detection schemes is also analyzed and compared.Finally,the simulation results show that the proposed two sensing schemes can achieve better sensing performance for non-Gaussian“Spectrum Shelter”signals,and are superior to the traditional energy detection schemes in resisting the uncertain noise.2.The IGS based covert communication scheme is studied.Via the above-mentioned “Spectrum Shelter”detection scheme,the appropriate authorized and public signals are selected as the“Spectrum Shelter”in real time to realize the transmission signal hiding.Firstly,the transmission rates of authorized users and cognitive covert users in the system are analyzed when cognitive covert users use IGS.Secondly,the influenceof fading characteristics of eavesdropping channel on the feasibility of covert com-munication is analyzed.Besides,the optimal detection threshold and the minimum error detection probability of eavesdropper are deduced.Next,the feasibility condi-tion for cognitive concealment users to select“Spectrum Shelter”is analyzed.Then,when the feasibility conditions are met,under the condition that the error detection probability of warden is higher than the preset value and the interference to the“Spec-trum Shelter”signal is not greater than the other preset value,the covert transmission rate is improved as far as possible through the joint optimization of the transmitting power and circularity coefficient of IGS.Finally,the simulation results verify that the proposed scheme can not only guarantee the covert performance,but also improve the reachable covert transmission rate.