Maximum Likelihood Optimization of Linear Frequency Modulated Signal Based on Particle Swarm

The maximum likelihood estimation (MLE) is an optimal parameter estimation for the linear frequency modulated (LFM) signal. It can reach the Cramer-Rao lower bound. However, the great calculation load caused by multidimensional search limits its practical application. In this paper, the particle swarm optimization (PSO) algorithms based on the MLE of LFM parameters are proposed. Three different PSO algorithms, namely global mode standard PSO, local mode standard PSO, and hybrid PSO combined with global mode and local mode, are applied to optimize the MLE of LFM parameters. Through the updating of velocities and positions of the particles in multidimensional space, the estimation speed of chirp rate and initial frequency of LFM signal is accelerated effectively. The convergence performance, statistical performance and calculation time of the three PSO algorithms are compared by MATLAB experiments, through which the performance of the three optimization algorithms for LFM parameter estimation is analyzed.