Community detection approaches are useful tools for revealing the structure and properties of networks. There are many approaches for identifying communities, some require inaccessible information, others need initializing parameters to perform well. Many existing algorithms require centralized decision maker to reveal communities. This paper proposes a non-centralized parameter-free method that only uses attainable information. The proposed method is examined using real as well as synthetic social networks. Experimental results suggest that this approach can perform as well as centralized approaches that require global information for detecting communities.