Abstract: For the optimization of spare linear array, a new method based on the improved adaptive particle swarm optimization (PSO) is proposed. Firstly, the modified PSO algorithm adjusts its inertia weight and learning factors adaptively via the adaptive policy according to the fitness of the particle, which improves the searching ability of the population, and then the velocity expression is modified，which guarantees the update of the velocity. To further accelerate the convergence rate, a crossover strategy is introduced when the algorithm is at a state of stagnation. This method completes the synthesis of the spare linear array with multi-constraint efficiently, and achieves lower peak sidelobe level (PSLL); numerical simulation shows the effectiveness of the algorithm.