Performance Analysis of Metaheuristic Approaches for Single-Snapshot DOA Estimation

This letter addresses the application of metaheuristics to the single-snapshot direction-of-arrival estimation. The performances of five of the most popular optimization algorithms-Particle Swarm Optimization, Ant Colony Optimization for Continuous Domains, Simulated Annealing, Genetic Algorithms, and Differential Evolution-are compared in terms of convergence, accuracy, resolution, and computational cost in order to find the most suitable one. Also, running parameters are suggested for all of them. The featured numerical simulations include both uniform and nonuniform arrays and different noise scenarios. Results show that all of the presented optimization algorithms are capable of solving the problem, but Ant Colony Optimization for Continuous Domains is the most recommendable.