A physical wireless parameter conversion sensor network (PhyC-SN) uses the frequency modulation of sensing information and recognizes the whole sensing information from the detected frequency spectrum. For long life sensor node, the event driven sensor decides to send the sensing information in accordance with the certain amount of changing about the sensing information. However, the data separation is difficult for the PhyC-SN with event driven sensor. This paper proposes the transmit control and the data separation for PhyC-SN with event driven sensor. The Gaussian probability model for predicting the future sensing information is used for estimating the difficulty of data separation. The effect of proposed technique is evaluated by computer simulation.