The line coding technology used in standard Ethernet networks and specified in the IEEE 802.3 standard. Manchester coding is the process by which digital information in a binary bit stream is converted into electrical signals for transmission.
Simple Manchester coding uses a two-state transition of line voltage to represent one bit of information. In other words, two baud (voltage changes) are used for one bit (piece of information). A binary 0 is represented by a transition from higher to lower voltage in the time set for transmitting one bit (that is, one “bit time”). A binary 1 is represented by a transition from lower to higher. For Ethernet networks, the high voltage is typically +0.85 volts and the low is typically -0.85 volts, making each voltage transition equal to 1.7 volts.
Manchester coding has the advantage of enabling data to be transmitted without the need for an extra clocking signal. This is possible because voltage transitions take place in the middle of each bit transmission interval, which establishes a timing pattern. The mid-interval voltage changes thus allow the sending and receiving stations to maintain proper synchronization with each other in order to ensure the integrity of the transmission. Because of the extra transition per bit that is used for clocking purposes, Manchester coding is only 50 percent efficient - for example, a 20-MHz bandwidth is required to produce a 10-Mbps data transmission rate.
Another version, called differential Manchester coding, represents binary 0 by a voltage transition at the start of the bit-interval, and binary 1 by no transition at the start of the bit-interval. In both cases, a transition takes place in the middle of the interval for synchronization purposes. Differential Manchester encoding is used for IEEE 802.5 token ring networking.
Graphic M-3. Simple Manchester coding for Ethernet.