The Research And Implementation Of Adaptive Beamforming Algorithm In Radar Signal Processing

At present,radar systems face complex electromagnetic environment and high performan-ce index requirements,which not only require real-time performance,but also require anti-jamming and anti-radiation properties.Digital beamforming is one of the core modules of the phased array radar signal processor,which can make full use of the received information of antenna array.Based on national ministry project,the research tasks of this thesis are mainly composed of two parts,adaptive beamforming algorithm and hardware implementation.1.Array data types and classical adaptive beamforming criteria are studied,and simple analysis and comparisons are made.The LMS?minimum mean square error?,RLS?recursive least squares algorithm?,and SMI?sample matrix inversion?algorithms are modeled and simulated,and the mathematical and physical properties of the three algorithms are compared and analyzed.For the complexity of the SMI operation,linear equations between the steering vector and the covariance matrix are established by introducing an auxiliary vector,and the linear equation is solved by using the CG?conjugate gradient algorithm?algorithm.Simulation analysis shows that the CG algorithm achieves the same results as the SMI by a finite number of iterations.In the practical application of radar,its beam dwell period is limited,so the number of data snapshots received by the array is also limited,so that the performance of the beamformer is degraded.In order to solve the problem,a diagonal loading algorithm is studied and combined with CG algorithm to obtain robust beamforming based on CG algorithm.The robustness of the algorithm under finite snapshots is verified by modeling simulation.2.For wideband adaptive beamforming,the ISM?non-coherent signal subspace method?and CSM?coherent signal subspace method?are compared and analyzed for non-coherent interference,coherent interference suppression,and the robustness for different frequency domain snapshots,where the CSM focus matrix is solved using the RSS?rotation signal subspace?algorithm.The simulation results show that compared with the ISM algorithm,the RSS algorithm has better robustness to changes in the number of frequency domain snapshots,and can suppress coherent interference.3.The design,function verification and synthesis of the narrowband beamformer hardware circuit are completed based on FPGA,and the CG algorithm is used to solve the weighting coefficient.Firstly,in the design part,according to the features that require multiple iterations in beamforming,the design is performed using floating point numbers;according to the structural characteristics of the CG algorithm,its key iterative modules are time-division multiplexed,reducing hardware overhead;for floating point numbers Aiming at the problems existing in floating-point accumulators when the number of delay cycles of floating-point adder is large,structural optimization is performed.Secondly,the verification platform is built by means of the joint simulation of Modelsim and Matlab,and the overall function is verified.The relative error between the MATLAB simulation results and the hardware circuit simulation results is in the order of 10^-5.This result is in line with the radar digital signal processing requirements.Finally,the FPGA synthesis of the beamformer is completed based on the Xilinx XC7VX690T device.The maximum clock frequency after synthesis is 185.481 MHz.