Ethernet Cables and Connection Types
The figures below illustrate what current standard RJ45 Ethernet sockets and plugs look like, and how an Ethernet cable connects.
RJ45 network cable, wall socket and connector illustration
 |  |
10BaseT and 100BaseT
The initial digits refer to the data transmission rate in megabits per second.
‘BASE' is for 'baseband' which describes the transmission of bits without further modulation, as distinct from 'broadband' which transmits multiple signals through ‘frequency division multiplexing'.
'T' is for Twisted Pair which is the kind of cabling used.
10BaseT carries 10 Megabits (not megaBYTES ) per second and was described by the original definition of Ethernet. This is carried on twisted-pair cable similar to phone cable called Category 3 cabling.
100BaseT, or Fast Ethernet, finalised in 1995, carries 100 Mbits per sec across Category 5 twisted pair cabling.
Gigabit Ethernet, finalised in 1998, carries 1000Mbits per Sec and can also be carried on Cat 5 cabling however Cat5e cabling is recommended.
Auto-Sensing: Most ethernet devices are auto-sensing, which means they automatically detect and adjust to the prevailing rate of data transmission.
CSMA/CD
As Ethernet is a ‘shared bus topology', every device connects via the same cable. So each device ‘listens' to the electrical activity on the cable and will only ‘speak' - send a frame of data - when every other device is silent. This technique is called Carrier Sense on Multiple Access Networks.
But collisions can still occur when two devices commence transmission at the same time. When this occurs, both the devices will detect the collision and wait for a random interval before trying again. The back-off algorithm which handles this process is called the Collision Detection, or CD, response.
Together, these methods make up CSMA/CD – Carrier Sense on Multiple Access Networks with Collision Detection. So, ethernet networks can be said to be divided into ‘collision domains' which are network segments shared by a number of computers.