Among existing DOA estimation methods, the Weighted Subspace Fitting (WSF) algorithm is well-known for its high resolution of DOA estimation. However, its computational complexity is extremely high and cannot meet the real-time requirements. In this paper, we propose a Joint-PSO algorithm for WSF with less complexity. This algorithm has the following key steps: firstly we use the solution of Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT) which can get the DOA estimation with extremely low complexity and stochastic Cramer-Rao bound (CRB) to determine a novel initialization space in the whole search space. Then, we randomly initiate a small number of particle in that small area. Finally, we let the particles “fly” to the solution with a suitable speed. Additionally, we also discuss and optimize the inertia factor of PSO algorithm. The simulation results find that for the same Root-Mean-Square-Error (RMSE), the particles and iteration number of the proposed algorithm are much less than that of the original PSO algorithm. As a result, the computational complexity can be greatly reduced.