The increase of P2P applications in campus networks has caused excessive consumption of bandwidth and network congestion, which reduces performance. This thesis uses the campus network as a background for research. It first analyzes the effect of P2P applications on the campus network. Next, it tries to resolve some of the present difficulties in P2P traffic identification by bringing forward a method for identifying P2P traffic with two major features, i.e. traffic and payload, using advantages of existing traffic identification technologies. Research shows that this method can effectively detect encrypted or unknown P2P applications with little cost and satisfying accuracy.