Abstract: With the WSN constrained by limited resources and energy for long, the deployment algorithm of sink nodes has become a key problem to WSN. However, the relevant existing algorithms have actually not been taking account of that the service provided by sink nodes often failed due to limited resources or the problems of wireless link. A deployment model of P-medians of multiple sink nodes is thus proposed with the computational solution given to the deployment model of the location theory and at the same time uses the genetic algorithm computing model of sink nodes which belongs to the NP-completeness problem. The computational accuracy and efficiency of the algorithm are also analyzed. The simulation results indicate that the deployment policy based on genetic algorithm can reduce the energy consumption of WSN efficiently so as to improve the efficiency of network service and prolong the network lifetime.