基于遗传算法的功率域最小二乘DOA估计器

提出一种基于遗传算法的功率域最小二乘波达方向(DOA)估计器。通过功率域最小二乘法推导得到优化目标函数。继而将目标函数作为适应度函数,将阵列天线各入射信号的未知DOA作为决策变量,通过遗传算法进行全局优化,从而完成各信源DOA的估计。采用适当的遗传策略,使全局寻优有效收敛至满意解,避免了复杂的多步迭代,提高了稳健性。估计器对非均匀高斯白噪声条件下的信源DOA同样有效。仿真实验表明,估计器的性能接近确定性最大似然(DML)估计器。

Power Domain Least Squares DOA Estimator Based on Genetic Algorithms

A power domain least squares direction of arrival(DOA)estimator was proposed.The optimization objective function was derived via the least squares method in power domain.With the objective function as fitness function and the unknown DOAs of all incident signals as decision variables,the global optimization was performed via the genetic algorithms(GA)to accomplish the DOAs estimation.The satisfactory solution for global optimization was obtained with proper genetic strategy,which avoided the complex iteration of multiple steps and improves the robustness.The proposed DOA estimator dealt well with not only the uniform white Gaussian noise but also the nonuniform one.Simulation experiments show the proposed estimator has an almost same performance as the deterministic maximum-likelihood(DML)estimator.