Membrane computing is a kind of biological heuristic computation，which is abstractedfrom the structure and function of the tissue or organ in living cells. Between two cells at most one linkis established，also known as synapses.Synapses have channel states associated with them. 'There-fore,based on this biological phenomenon， a new type of computing model is proposed : A communi-cation membrane system with channel states.Communication P systems with channel states are a classof distributed parallel computing models，the communication of this kind of P system mainly dependson the symport/ antiport rules，where the rules on each channel are executed in a sequential manner,and the state on the channel is used to control the communication between two cells or between a celland the environment. 'This work summarizes the conception of cell-like communication P systems,tis-sue-like communication P systems，cell-like recognizer communication P systems，tissue-like recog-nizer communication P systems according to the symport/ antiport rule with channel states，and thecomputational power and computational complexity of this type of system are explained. Finally，asummary and a future outlook are given.