Abstract: Network congestion has become a widespread concern. At present, research based on network congestion control has present some specific methods and suggestions, but they arc not able to solve the network congestion well.Qualitative Dynamic Probabilistic Networks(QDPNs) are one of the most efficient models in the uncertain knowledge and dynamically reasoning field. This paper presented a network congestion control method based on QDPNs. This method can effectively predict network congestion, and then we can get the corresponding congestion control strategy according to the prediction before.

Key words: Qualitative dynamic probabilistic networks,Network congestion control