| With the rapid increasing number of mobile communication users and the increasing demand for wireless communication data transmission rate and system capacity,the spectrum resources of low-frequency can no longer meet people's needs,while millimeter wave attracts people's attention with its rich bandwidth resources.The combination of millimeter wave and massive MIMO-OFDM technology can solve the serious path loss and selective fading problem in millimeter wave communication,and improve the system capacity,spectrum utilization,and anti-interference ability.In order to implement the directional power gain of the large-scale antenna array,the transceiver needs to use the channel state information to design beamforming algorithms and precoding matrixs,but due to the sparse scattering nature of the millimeter wave channel and the large number of large-scale antennas,the complexity of channel estimation is increased.At the same time,in the uplink pilot training transmission,when a pilot spoofing attack occurs,it will cause the base station wrongly estimate the channel state information of the user,and the information between the base station and the legitimate user will be easily leaked during the downlink data transmission.In view of the above situation,this thesis mainly studies the secure channel estimation problem of the millimeter wave massive MIMO-OFDM communication system.The main research results are as follows:(1)First,for the wideband millimeter wave communication system with frequency selective fading,proposed a decomposable three-dimensional atomic norm minimization(ANM)method for uplink channel estimation.By exploiting the sparse characteristics of the millimeter wave channel and the channel matrix is composed of the superposition of the sine vectors of the channel parameters(angles of departure/arrival and time delay),the channel estimation problem can be regarded as a three dimensional line spectrum estimation problem.Using the three dimensional ANM algorithm to solve this channel estimation problem can obtain high-resolution channel parameters and the channel matrix.In order to reduce the computational complexity of the three dimensional ANM channel estimation algorithm,the channel estimation is divided into two steps.Firstly,the decoupled two dimensional ANM is used to solve the channel estimation problem of the millimeter-wave massive MIMO system.Secondly,the one dimensional ANM is used to solve the channel estimation problem of the OFDM system.Simulation results show that our proposed algorithm is better than other algorithms in estimation accuracy,and the computational complexity of our proposed algorithm is significantly reduced compared with the three dimensional ANM.(2)Secondly,for pilot spoofing attacks generated by eavesdropper in communication systems,proposed a multi-path channel detection algorithm and the channel state information recovery solutions.This paper presents a detection algorithm based on the random matrix theory,this algorithm determines the existence of pilot spoofing attacks by analyzing the distribution characteristics of eigenvalues in the noise subspace of the received signal.When the detection algorithm get the conclusion that the pilot spoofing attack exists,in order to eliminate the damage caused by the pilot spoofing attack,this paper proposes two channel parameter predicted methods based on the Autoregressive Integrated Moving Average model and the Long Short Term Memory model,these methods can effectively predict the current channel state information according to the historical value of the channel parameters.Simulation results show that the proposed pilot spoofing detection algorithm can achieve accurate detection even under bad signal noise ratio conditions,and its performance is better than other detection algorithms;the two proposed channel parameter prediction methods can effectively recover the channels matrix in a short time.This thesis consists of 5 chapters,including 25 figures,9 tables and 67 references. |
PDF Full Text Request