10 April 2007


Developed by Bob Metcalfe at Xerox's Palo Alto Research Center (PARC) in the 1970's for supporting local area networks (LANs), Ethernet was part of a large raft of information technologies that emerged that year.* It is both a cable/socket design, and a digital data transmission format.

The most familiar form is the "fat telephone" or 8-wire RJ45 cable (see figure). There is also something called "industrial ethernet," which are really just extremely high bandwidth (>10 megabits per second-Mbps) Internet connections, not true Ethernet at all.

In order to understand the concept of Ethernet, it's necessary to mention the 7-layer OSI Reference Model.

The OSI Reference Model
(OSI article)

The OSI specifies how data from an application in one computer moves through a network medium to an application running on another computer. The 7 layers are organized as media(1-physical, 2-data link, 3-network, 4-transport) and host(5-session, 6-presentation, 7-application). Media, here, refers to the mode by which data is transmitted over distance; e.g., over microwaves, coaxial cable, and so on. Host, here, refers to any computer on a network that is capable of running applications. All devices on a network, regardless of capability or description, are called nodes. (Nowadays, cases of the node that can't run a program are rare; they include dumb terminals). So the "host layers" are responsible for the final communications between the network and the application (such as a web browser).

The physical layer includes the specs for cables and routers. The data link layer consists of the precise format for encoding data for transmission. The data layer for the TCP/IP suite is where the Ethernet format resides. The connection between the two formats lies in the signal frequency. Ethernet requires cables that support a range of frequencies, which creates very demanding performance requirements for cable and especially connectors. The interaction between Ethernet (the data format) and the network format takes the form of Ethernet II Framing. In the Ethernet format, each packet is an IEEE 802.3 frame, which specifies the format, content, and size of each packet. Each frame consists of the destination and originating MAC address (6 bytes each), followed by a 2-byte sequence that gives the length of the frame. The frame has to be between 64 and 1518 bytes in length. There is also a frame check sequence (FCS) that is used to check the integrity of the frame; this occurs at the end of the frame.

Inside that package, there is a logical link control (LLC) that communicates to the network interface card (NIC) the memory buffer from where the frame came, and where it goes. This allows the same NIC to support multiple network protocols concurrently. Frequently LLC and MAC addresses are regarded as two different sub-layers within level 2 [*].

The next layer up is the network layer ("Internet layer"); this is the format that describes the actual network structure. In addition to the Internet Protocol (IP) , there are also several mostly-defunct networking protocols like SNA, token ring, and so on. The difference between networking protocols, in the least abstract sense, is the routing decision. In IP, routing goes "outbound" to a router, then to another router, and eventually to the correct host. In a token ring (to take an alternative concept), a packet is routed to the next host on the "ring," along with the "token."

IP is the component of the Internet suite that resides at the network layer (3); it reads the MAC address and uses it to route the packet in accordance with the network architecture. Part of the technical difficulty posed by IP is that it addresses a network of networks. Hence, a frame is likely to pass through many routers as it goes from network to network. The part of the network that an individual router is connected to is called a subnet; each node in the network is connected to a router within the subnet. The level 3 IP address directs the packet/frame to the correct subnet, but requires the address resolution protocol (ARP) to convert the IP address to the correct (layer 2) MAC address.

Peculiarities of Ethernet
Ethernet is an architecture used for local area networks (LANs). The physical component of Ethernet—its radio frequency (RF) transmissions and high signal density meant that data loss and distortion became a serious problem if network segments exceeded 75 meters. That limitation has been pushed back to tens of kilometers, so that college campuses and even major cities may be connected by the extremely high bandwidth LANs Ethernet allows. At the same time, transmission rates for Ethernet are now approaching 10 billion bits per second (Gbps).

Ethernet allows the transmission of data along circuits with many nodes on each circuit. This means that all the different nodes on a circuit will receive all of the data transmitted down the circuit, but each NIC will be able to select only the data addressed (illustration). Another important aspect of the Ethernet protocol is that it is decentralized; hence, all hosts on the network must follow uniform rules such that there is no crosstalk (CSMA/CD protocol). This allows Ethernet to support a very large number of possible network architectures.
* Another was the C programming language; in '73, the Unix kernel was rewritten in C, which may be said to be the true birthday of that operating system. Also, the TCP/IP specification was developed and formalized that year, beginning the Internet. Yet another event was the release—also by Xerox PARC—of the Alto, which pioneered the graphical user interface (GUI) and the mouse.

Cables: Ethernet and "industrial ethernet" (notice the capitalization) have been routed over several types of cables. As I've mentioned, the most familiar is the RJ45 (EIA/TIA Category 5E twisted four-pair) ,which looks like a fat telephone jack. This has 8 wires, compared to the 2 in a phone jack.

Another cable design is the coaxial cable most commonly used for cable television. Ethernet versions of this are distinguished by the slightly different connector; while coaxial TV cable has a C connector, cable Ethernet uses the BNC connector. Usage has greatly diminished in recent years. (SOURCE: Edward F. Kuester, "Common Coaxial Connectors") The BNC/coaxial system is similar in appearance to the M12 connector, which is being marketed as a new industrial ethernet format.

A third format is the SC/SC Duplex, a type of fiber optic industrial ethernet connector. Industrial ethernet standards such as these are semi-standard or proprietary extensions of the Ethernet (IEEE 802.3) standard; they are intended to allow very high-bandwidth (>10 million bits per second) LAN media in an environment that may be very harsh to electronic components. (SOURCE: Emcore, Inc., " Fiber Optic Connectors")

Routers: a computer networking device that links nodes in a network, and routes transmissions among them. Routers vary immensely base on the networking format used and the physical mode (coaxial, fiber optic, RJ 4-pair twisted) used; however, all routers do basically the same job. Based on the data supplied by the frame, the router can identify where it must logically dispatch each data packet. Through a succession of logical branches, the frame will eventually reach the correct destination. (SOURCE: Wikipedia)
ADDITIONAL READING & SOURCES: Howstuffworks, "How Ethernet Works"; AutomationWorld, "Getting physical with industrial ethernet"; Cisco Systems, "Internetworking Basics"; very odd but entertaining explanation may be found at RouterGod ,"Robert Downey Jr. on the IEEE 802.3 Ethernet Frame"—this was the one coherent explanation I could find that incorporated a reference to a voluptuous sheriff's deputy and scented body lotions. See also "CCNA Bootcamp: the OSI Model";

Wikipedia: OSI Reference Model, Ethernet, Ethernet II Framing;



Post a Comment

<< Home