Study On The Physical Layer Security Technology Of Ultra-Dense Network

With the exponential growth of mobile data and the popularity of traffic-intensive applications,people's requirements for existing wireless communication networks are getting higher and higher,and the increasingly scarce spectrum resources can no longer meet the demand for high-quality communication services.For this reason,in the 5G,a User-centric Ultra-Dense Networks(UUDN)architecture was proposed.UUDN provides users with the best network services by deploying a large number of small access nodes in hotspots and organizing Access Points Group(APG)around specific users,which is the mainstream trend of the future 5G network architecture.The increasing number of access nodes and the shrinking distance between nodes and users make the wireless communication environment of UUDN system more complex and changeable,which makes the data transmission seriously threatened.The physical layer security technology uses the uniqueness,reciprocity,and time-varying characteristics of the wireless channel to encrypt data.It neither involves complex cryptographic calculations nor excessively requires the computing power of the communication entity,which can significantly reduce the burden on small access nodes and mobile terminals,so it is an important research direction for data secure transmission in the UUDN system.Aiming at the active and passive eavesdropping situations in the UUDN system,this paper proposes an active eavesdropping detection algorithm and a passive eavesdropping interference strategy using physical layer security technology.The main work and innovations of the paper are summarized as follows:(1)Aiming at the active eavesdropping scene in the UUDN system,a multi-node joint detection algorithm based on the LS-FDC criterion is proposed.First,the linear shrinkage(LS)theory in statistics is used to shrink and optimize the covariance matrix of the sample to better fit the distribution of the overall eigenvalues after eigen decomposition.Then,each node in APG uses the Flexible Detection Criterion(FDC)algorithm to jointly determine whether there is an active eavesdropper.Simulation experiments and theoretical analysis show that:compared with the traditional Minimum Description Length(MDL)algorithm and FDC algorithm,when the pilot length is much larger than the number of node antennas,the detection probability of this algorithm at low signal-to-noise ratio is significantly improved;When the pilot length is less than the number of node antennas,the detection capabilities of other algorithms are severely invalid,and the detection probability of this algorithm can still maintain a high level.(2)Aiming at the passive eavesdropping scene in the UUDN system,a secure downlink data transmission scheme based on artificial noise assistance is proposed.First,when the sending node transmits downlink data,the beamforming and artificial noise joint technology is used to add a piece of artificial noise to the zero space of the transmitted signal to ensure that it does not affect legitimate users while increasing the interference of the eavesdropping channel,thereby achieving secure information transmission.Then,aiming at the problem of artificial noise and useful signal power allocation,an optimal power allocation strategy that satisfies the legal user's Quality of Service(QoS)and the minimum power allocation coefficient is given.This allocation strategy has been optimized between performance and node power,making it more suitable for the transmission scenario of the UUDN system:Under the constraints of user QoS and the lowest power allocation coefficient,in order to meet the needs of downlink transmission,the nodes in the APG will give priority to increasing their transmission power.If the constraint conditions are met within the maximum power threshold of the node,data transmission is allowed;otherwise,data transmission is prohibited.Finally,combined with simulation,the influence of factors such as the number of node antennas,signal transmission distance,and received signal-to-noise ratio on the power allocation strategy is analyzed.