Ethernet technology provides rules that allow network-connected devices to talk to one another without talking over each other. In a verbal conversation, when two people speak at the same time, each may have difficulty understanding what the other is saying. This is amplified when, say, 10 people are talking at once. Imagine 100 or 1000 at once.
The same is true for a data network. If two or more connected devices on a shared network attempt to transmit data packets at the same time, a packet collision occurs. The pulses of electricity or photons that make up a packet overlap when sent at the same time over a shared copper or optical cable. This jumbles the sequence of "on" pulses and "off" voids that are sent to indicate the bits and bytes of 1s and 0s that make up a packet.
Ethernet was designed to solve the problem of packet collision. It provides network devices with a set of rules that essentially says: "Make sure no one else is talking before you talk. If you hear someone talking while you're talking, stop, listen, and wait for the talking to end before you talk again."
Ethernet is commonly associated with connected devices in a wired LAN or WAN. Using a wired Ethernet cable, devices are connected to an Ethernet switch. Ethernet has the ability to use both wired and fiber cables, delivering not only data but also power, now up to 90W with UPOE+.
More formally, Ethernet is a common name for the IEEE 802.3 standard based on the Carrier Sense Multiple Access/Collision Detection (CSMA/CD) protocol. CSMA/CD defines when to transmit and what is to happen if a collision is detected, as well as endpoint addressing, transmission speeds, and media. Ethernet has evolved dramatically since its first application. Today it's the de facto protocol for IP-based networks.
The concept of Ethernet has its roots in the late 1960s and the University of Hawaii's Aloha Network. Aloha was a pioneering, experimental radio-communications network connecting the Hawaiian Islands to a central time-shared computer on the main campus, in Oahu.
Aloha was often referred to as one of the first wireless packet networks. It used two radio frequencies, separating send and receive data that passed between user terminals and the main hub connected to the computer. Designed for simplicity, the network followed these rules:
As use of the network grew, it became obvious that packet collision would severely limit the capacity of the network as designed. Researching the problem for his doctoral thesis, a young computer science student, Bob Metcalfe, devised a solution. His innovation earned him not only his Harvard Ph.D. but a place in history as the inventor of Ethernet. Metcalfe's solution: Listen before you talk.
In 1972, Metcalfe took Ethernet from idea to fruition with the first experimental Ethernet system, created at Xerox PARC. The network had an initial speed of 2.94 Mbps and linked Xerox Altos mainframes, servers, and printers using inexpensive coaxial cable.
At the time, Ethernet competed with Token Ring, ARCnet, Fiber Distributed Data Interface (FDDI), and other proprietary protocols. It quickly rose to dominance as it was open and reliable, yet inexpensive. Today, it supports speeds from 1 Mbps to 400 Gbps and beyond.
Ethernet was formalized in 1983 as the IEEE 802.3 Ethernet standard. It is one of 12 in the IEEE 802 family of variable-size packet LAN standards. IEEE 802 services and protocols focus on the physical and data link layers of a network. In the Open Systems Interconnection (OSI) networking model, those layers are known as Layer 1 and Layer 2, respectively.
Based on the addressing scheme used in early implementations of Ethernet, the IEEE 802 standards use Media Access Control (MAC) addresses, which are unique identifiers typically assigned to endpoint access hardware by device manufacturers. The most significant part of a MAC address identifies the manufacturer, which assigns the remainder of the address, potentially providing a unique address. This makes it possible for frames to be delivered within a network that interconnects sender and receiver through some combination of Ethernet switches.
Routers are used to pass packets between networks and use IP addresses rather than MAC addresses to tag packets' origination and destination points. Using the U.S. Postal Service as an analogy, IP addresses are like ZIP codes, while MAC addresses are like street addresses or post office boxes.
In an Ethernet network, data is broken into packets, with each packet transmitted using the CSMA/CD algorithm until it arrives at its destination without colliding with any other packet. The first open slot after a transmission is reserved for an acknowledge packet. A device or node is either transmitting or receiving at any instant.
When sending data to another device on an Ethernet network, the MAC sublayer (data link sublayer):
When receiving data from the physical layer, the frame check sequence in the MAC block is used to ensure data integrity. It strips off the sender's Ethernet-packet preamble and padding before passing the data to the higher layers.
The Ethernet switch creates the frame by encapsulating the Ethernet packet with a preamble, which is used to synchronize the sender and receiver, followed by a 1-octet start-frame delimiter byte.
The Ethernet packet includes the data encapsulated by a header and a frame check value. The MAC header provides the source and destination MAC addresses as well as a 2-octet EtherType code indicating IPv4 or IPv6, MACsec encryption, etc. The frame ends with a frame check sequence (FCS), which is a 32-bit cyclic redundancy check used to detect any in-transit corruption of data.
When a frame is ready, the transmitting station checks to see whether the channel is idle or busy. If the channel is busy, the station waits until the channel becomes idle. If the channel is idle, the station starts transmitting and continually monitors the channel to detect collision.