In step 4 of the CSMA/CA protocol, a station that successfully transmits a frame begins the CSMA/CA protocol for a second frame at step 2, rather than at step 1. What rationale might the designers of CSMA/CA have had in mind by having such a station not transmit the second frame immediately (if the channel is sensed idle)?
If the channel is detected idle then the station does not transfer the second frame to avoid the collision
by design of the CSMA/CA. When received the protocol for first frame, a station to send second frame
begins at step 2 rather than step1.
For example:
Station X and Station Y are transfers data frames,remain all not want to transmit data. In this situation,
station X find the channel to be idle and send first frame then next second frame. Similarly Y received
the transfer data at the same time. Then leads to collision. So, using CSMA/CD to avoid collsion.