Distributed Pareto optimization via diffusion strategies