Research On Null Space Projection Beamforming For Full Duplex Self Interference Cancellation

In recent years,with the development of communication technology and the continuous enrichment of information services,people’s demand for wireless communication is also increasing,and the most direct way to increase the transmission rate of wireless communication systems is to improve the spectral efficiency and transmission bandwidth of the system.The full-duplex millimeter wave system realizes the improvement of spectral efficiency and transmission bandwidth.Millimeter wave communication provides a wider frequency band.Full-duplex technology can transmission and receive signals at the same frequency at the same time,making full use of the wide frequency band of millimeter wave and obtaining twice the ideal spectrum utilization.However,to achieve full-duplex technology,self-interfering signals must be resolved.This paper mainly studies the problem of self-interference cancellation based on null space projection.Firstly,this paper analyzes and compares the performance of two self-interference cancellation methods.One of them is based on null space projection,and the other is based on self-interference(SI)reconstruction.In ideal channel information,the total information rate of the two methods is derived.In the imperfect channel information,the expected SI energy and the approximate ratio of expected signal to self-interference plus noise rate(SINR)are derived.The results show that the method based on null space projection can suppress SI more effectively under the condition of imperfect channel information.Secondly,the theoretical analysis of the transmit beam mainly analyzes the communication coding error and the interference suppression ratio.The null-space projection matrix formed by self-interference channel through eigenvalue decomposition is introduced.Self-interfering channels are considered with additive channel errors.Assuming that the ideal communication coding is known,the actual transmission precoding is designed and the communication coding error is analyzed.Then,the interference suppression ratio is analyzed in detail,and the trade-off curve between coding error and interference suppression ratio is established according to the subspace dimension.Thirdly,an improved analog encoding is proposed.A two-dimensional plane is established based on millimeter-wave angle information,and an orthogonal analog code is designed.However,due to the limitation of the resolution of quantization pairs on the twodimensional plane,some selected quantization pairs cannot be located in the communication space.An improved scheme based on space boundaries is proposed.By bounding the communication space in a rectangular space,an improved quantized pair is established.Then,with the help of circular search,the quantization pair farthest away from the SI region is found,and an improved analog coding matrix is calculated to realize the elimination of self-interference.Finally,in the design of orthogonal analog codes,it is assumed that the Light of Sight(LOS)path of SI does not exist.Therefore,an analog code design based on perturbation is proposed.A perturbation is added to the quantization pair corresponding to the orthogonal analog code to create a new quantization pair candidate set.Three schemes for solving perturbation,independent block,joint block and based on particle swarm optimization algorithm,are proposed.Simulation experiments show that the perturbation-based method can suppress 10 dB more LOS when the antenna isolation is low.