Stored in an each speaker's sound volume coefficient table 13 are a speaker identification ID and information of a voice coefficient so as to be correlated with each other. A speaker identification unit 11 extracts an each speaker's ID from a received packet based on floor control information or outputs the packet as an RTP packet to an RTP buffer. A sound volume control unit 12 executes a control program 14a of a recording medium 14 to record a current speaker identification ID and reads a sound volume coefficient relevant to the speaker identification ID from the each speaker's sound volume coefficient table 13 to output the coefficient as sound volume control information. In this case, the sound volume control unit 12 increments/decrements the relevant sound volume coefficient based on sound volume adjustment information of a sound volume adjustment unit.