Research On Key Techniques Of Physical Layer Security In Wireless Network

In the fast-changing information age,more and more people rely on the use of wireless networks(such as cellular networks and Wi-Fi)for the exchange of important information,such as credit card information,financial data and personal mail.However,due to the openness and broadcasting nature of wireless media,wireless communication systems are vulnerable to eavesdropping attacks.In recent years,the global cybercrime has become more and more intense,which makes the security aspects in wireless communication receive much attention.Traditionally,the security of wireless systems relies on the encryption technology implemented at the upper layer.However,its performance depends on the complexity of computation.Moreover,the encryption algorithm can be cracked by exhaustive attacks,which can not achieve perfect secure communication.Accordingly,the scholars in the field of communication turn their eyes to the physical layer.Unlike cryptography-based encryption technology,physical layer security makes use of the inherent random characteristics of wireless channels to achieve secure transmission of information.It does not require key distribution and management,and does not restrict the computing power of eavesdroppers.It has become a research hotspot in the field of wireless communication.In order to meet the increasing demand for security in wireless communication networks,based on different level of eavesdropper's channel state information(CSI)available,this thesis provides security guarantee for different wireless communication systems by exploiting cooperative jamming,artificial noise(AN),relay selection and power allocation.Combining with information theory,probability theory and convex optimization theory,the security performance of the system is also analyzed.The specific research contents and objectives of this thesis are as follows:Firstly,the current research on physical layer security performance analysis is mainly restricted to a specific fading channel(such as Rayleigh channel,Rice channel,etc.),which is only applicable to specific wireless communication application scenarios.However,with the emergence of a large number of new scenarios in 5G networks,such as Internet of Things(IOT),Device-to-Device(D2D),and so on,these traditional statistical distribution models have been proved unable to match the experimental data of these new scenarios.To better model these emerging scenarios,Yacoub recently proposed a more generalized?-?-?-? statistical distribution model,which comprises of almost all the fading models in the existing public literature.However,no work has been done to analyze the physical layer security performance based on this channel model.To address this problem,this thesis analyses the physical layer security performance of classical Wyner wiretap model based on ?-?-?-? fading channel.Specifically,the analytical expressions of average secrecy capacity(ASC),asymptotic ASC,secrecy outage probability(SOP)and the probability of non-zero security capacity are derived.Secondly,this thesis studies a multiple-input multiple-output multi-antenna eavesdropper(MISOME)underlay cognitive wireless network,in which one multiple-antenna secondary transmitter intends to achieves secure communication with a single antenna secondary receiver,and be granted to share the licensed spectrum of one primary user,while a passive multi-antenna secondary user attempt to overhear the secondary transmission.With the eavesdropper's CSI is completely unknown,this thesis proposes an AN-assisted beamforming secure transmission scheme to improve its physical layer security performance,and studies the power allocation scheme based on SOP minimization.With at least the statistical distribution information of the eavesdropper,the existing power allocation schemes derives the expression of SOP by fixing the power allocation factor,and then obtains the optimal solution via using one-dimensional linear search.However,no analytical solution is obtained.Moreover,the accuracy of this algorithm depends on the search interval,which increases the complexity of the system.To address this problem,this thesis proposes an adaptive power allocation scheme,derives the analytical solution of power allocation factor,and derives the closed-form analytical expression of SOP on this basis.Thirdly,in the DF cooperative relay network coexisting with a single eavesdropper,this thesis proposes a joint relay selection and cooperative jamming two-stage secure transmission scheme.In the first stage,the source sends confidential information to the pre-selected forwarding relay,while the friendly jammer and the destination send jamming signals to cover the transmission of confidential information without affecting the forwarding relay.In the second stage,the forwarding relay decodes and forwards the signal,and jointly transmits the jamming signal to reduce the reception quality of the eavesdropper without affecting the destination.The problem of power allocation between information signals and jamming signals based on secrecy rate maximization(SRM)under a total power constraint and total transmission power minimization under a achievablesecurity rate constraint is studied.The optimal solution is obtained using a combination of a bisection method and a one-dimensional search.In order to further enhance the system's achievable secrecy rate,this thesis proposes the optimal relay selection scheme and the suboptimal selection scheme.The theoretical analysis and simulation results show that the suboptimal relay selection scheme can achieve the near-optimal performance when the number of relays is slightly large.Fourthly,this thesis studies the secure secondary transmission in an underlay cognitive cooperative relay network coexisting with one multi-antenna eavesdropper or one single-antenna eavesdropper,and proposes a joint relay selection and cooperative jamming secure transmission scheme to improve the secrecy performance of secondary network.To be specific,one of the cognitive relays is optimally selected from the successful decoding set to retransmit the information signal,while the remaining ones emit jamming signals to confuse the eavesdropper without affecting the reception quality of the legitimate receiver.The existing relay selection schemes are either based on the instantaneous CSI of eavesdropper,or they do not make full use of the CSI of interference link.To address this problem,with only the channel distribution information of the eavesdropper,this thesis proposes two relay selection schemes,named conventional relay selection(CRS)and best relay selection(BRS)schemes,to further improve the secrecy performance for secondary network.In the CRS scheme,only the CSI of the main channel is employed,whereas the BRS scheme combines both the CSI of the main channel and interference channel.The closed-form analytical expressions of SOP under CRS and BRS schemes are derived.The simulation results under the above two scenarios show that the performance of the BRS scheme significantly outperforms than that of CRS one.