In recent years, most of the existing researches on mobile cloud computing focus on offloading compute-intensive tasks from mobile devices to remote cloud to save energy for mobile devices. But the longer transmission distance between mobile users and remote clouds will lead to the long latency of access. If we deploy some cloudlets nearby mobile devices to execute the offloaded tasks form the devices, then this problem can be solved. In this paper, we focuse on the cloudlet placement problem in the wireless metropolitan area network. By deeply studying the characteristics and corresponding principles of the spectrum clustering algorithm, combined with the actual needs of mobile mobile users in the wireless metropolitan area network, the corresponding cloudlet placement model is proposed. The corresponding cloudlet placement algorithm is designed to find the optimal partition and micro cloud placement location to minimize the average access delay of mobile users to the cloudlet in the region. Finally, the performance of the proposed algorithm is evaluated by experimental simulation. The results show that the algorithm can effectively reduce the average access delay of the cloudlet.