This paper proposes a framework of particle swarm optimization (PSO) based pairwise testing. To systematically build pairwise test suites, two different PSO based strategies are proposed. One strategy takes on a one-test-at-a-time approach and the other takes on an IPO-like approach. In these two different strategies, PSO is used to complete the construction of a single test and research on how to formulate the search space, define the fitness function, and set some heuristic settings. To verify the effectiveness of this approach, these algorithms are implemented and some typical instances have been chosen. In this empirical study, the paper analyzes the impact factors of this framework and compares this approach to other well-known approaches in test suite size and generation time. Final empirical results show the competitiveness of this approach.