Aimed to the problem of unified resource optimized allocation in executing multi-virtual network embedding synchronously, a multi-objective particle swarm optimization based virtual network embedding method is proposed to increase the performance of resource utilization and global load balance of the substrate network resource. A multi-objective optimization model for virtual network embedding was put forward, on the basis of which the single virtual network embedding is considered as a swarm and the information exchange among different swarms so as to evolve corporately by employing multi-swarm coordination optimization method, and finally the increase of the performance of global resource allocation was achieved. The simulation results show that the proposed method maintains the best performance of the substrate network resource utilization ratio and virtual network request acceptance ratio compared with typical methods. 