Physical Layer Security Theory And Methods Based On Degrees Of Freedom Analysis

The area of physical(PHY)layer security has been receiving a lot of attention for decends,from the point of view of secrecy rate,i.e.,the rate at which the legitimate receiver can correctly decode the source message,while the eavesdropper obtains no useful information about the source signal.Existing results show that multi-antenna as well as artificial-noise based techniques are promising methods to improve the performance of secrecy.However,those techniques have been facing a number of significant problems.1)Determining the maximum achievable secrecy rate in this scenario generally requires solving a nonconvex secrecy rate maximization problem,and so it is still an open problem.2)The newly introduced noise signal input covariance matrix is coupled with the signal input covariance matrix in the terms of the achievable secrecy rate,which makes the determination of the maximum achievable secrecy rate more difficult.3)In more complicated application scenarios,such as the co-existing of secrecy communication networks and public communication networks,the bi-directional communication networks and the active-evesdropper networks,the uncover of the relationship between the secrecy performance and the system parameters becomes more difficult.The study of secure schemes with high efficiency?robustness and low computational complexity,is a meaningful task in practical systems.The specific works of this dissertation include:1).As an alternative to secrecy rate,we propose to use the achievable secrecy degrees of freedoms(SDo F)as the performance metric.The SDo F is defined as the rate at which the secrecy rate scales with log(P)for the high SNR regimes.As compared with the achievable secrecy rate,the determination of the SDo F is a simpler problem.The SDo F results enables us gain more insight into how the maximum achievable secrecy rate behaves with the number of antennas at each terminal.2).For the purpose of solving the SDo F maximization problem,we first propose a secrecy cooperative secrecy transmission scheme under the scenario of a helper-assisted Gaussian wiretap channel with a source,a legitimate receiver,an eavesdropper and an external helper,where each terminal is equipped with multiple antennas.In that proposed scheme,the subspace spanned by the message signal has no intersection with the subspace spanned by the artificial nosie signal at the legitimate receiver,and belongs to the subspace spanned by the interference signal at the eavesdropper.Subsequently,we prove that such cooperative scheme is sufficient to achieve the maximum SDo F.In this way,we reduce the number of precoding matrices we need to investigate in.Based on this cooperative secrecy transmission scheme,we then determine the maximum achieveable SDo F result in closed form,and also provide the precoding matrices which achieve the maximum SDo F in closed form.3).We extend the aforementioned results to the case of MIMO two-user wiretap network,i.e.,a source destination pair exchanging confidential messages,another pair exchanging public messages,and a passive eavesdropper who is interested in the communications of the former pair.The maximum achievable SDo F region boundary points are obtained in closed form,and the construction of the precoding matrices achieving the maximum SDo F region boundary is provided.The proposed expressions are functions of the number of antennas at each terminal,and apply to any number of antennas,thus constituting advancement over prior works that have considered only fixed antenna configurations.Numerical results show that the network performance benefits when the two users get closer.As expected,the secrecy rate degradation due to channel state information estimate error can be counteracted by bringing the two users closer together.4).We further consider linear precoder design for a MIMO Gaussian wiretap channel,which comprises two legitimate nodes,i.e.,Alice and Bob,operating in Full-Duplex(FD)mode and exchanging confidential messages in the presence of a passive eavesdropper.The maximum achievable sum SDo F is given in closed form.The method for constructing the precoding matrix pair which achieves the maximum sum SDo F is provided.Numerical results show that,apart from the advantage of higher spectral efficiency,the FD based network also provides a good structure in terms of keeping messages secret.Also,the proposed secrecy transmission scheme is robust to self-interference and the conventional vulnerable positions of the eavesdropper.5).We consider a scenario in which an Alice-Bob pair wishes to communicate in secret in the presence of an active Eve,who is capable of jamming as well as eavesdropping in FD mode.As countermeasure,Bob also operates in FD mode,using a subset of its antennas to act as receiver,and the remaining antennas to act as jammer and transmit noise.With a goal to maximize the achievable SDo F of the system,we provide the optimal receive/transmit antennas allocation at Bob,based on which we determine in closed form the maximum achievable SDo F.We further investigate the adverse scenario in which Eve knows Bob's transmission strategy and optimizes its transmit/receive antennas allocation in order to minimize the achievable SDo F.For that case we find the Worst-case achievable SDo F.Our analysis has revealed that a positive SDo F can be guaranteed as long as Bob is equipped with more antennas than Eve.Numerical results validate the theoretical findings and demonstrate the performance of the proposed method in realistic settings.