Before switches, networks based on hubs data could only allow transmission in one direction at a time, this was called half-duplex.
By using a switch this restriction is removed; full-duplex communication is maintained and the network is collision free.
Fully switched networks employ either twisted-pair or fiber-optic cabling, both of which use separate conductors for sending and receiving data.
[3] In this type of environment, Ethernet nodes can forgo the collision detection process and transmit at will, since they are the only potential devices that can access the medium.
The core purpose of a switch is to decongest network flow to the workstations so that the connections can transmit more effectively; receiving transmissions that were only specific to their network address.