Index


Local area networks page

    General information

    A typical network consists of nodes (computers), a connecting medium (wired or wireless), and specialized network equipment like routers or hubs. In the case of the Internet, all these pieces working together allow your computer to send information to another computer that could be on the other side of the world!Typical networks today use two differentaddressing mechanisms on top of each other - and addressing is anecessity for data exchange between any two networked machines.The lowest level addressing is the MAC acess (akaethernet addresses, hardware addresses). The MAC addresses are used foraddressing within a single LAN. MAC addresses are programmed intothe hardware (typically network adapters), Every Ethernet card hasan unique MAC address (it is possible to change MAC on most adapters it'spossible, but not advisable except in special circumstances).The next address level are IP (Internet Protocol) addresses. These arein the form of "192.168.105.1" (four dot-separated numbers). An IP address is not programmed into hardware, but is set bysoftware to either a fixed value for a machine, or can be queried froma server somewhere in the local LAN. Also other addresses can be used if other protocols than IPare run on the LAN system. Some other protocols which areused sometimes in LANs are IPX, NetBIOS, DECnet, Banyan, etc.

    • Bridging and Switching Basics - Bridges and switches are data communications devices that operate principally at Layer 2 of the OSI reference model. As such, they are widely referred to as data link layer devices.    Rate this link
    • IEEE 802.1P - The IEEE 802.1P signaling technique is an IEEE endorsed specification for prioritizing network traffic at the data-link/MAC sublayer (OSI Reference Model Layer 2). The 802.1P standard also offers provisions to filter multicast traffic to ensure it does not proliferate over layer 2-switched networks. The 802.1P header includes a three-bit field for prioritization, which allows packets to be grouped into various traffic classes.    Rate this link
    • Home and Small Office Networking with Windows XP - Is a home or small office network right for you? There are lots of reasons to consider one. For example, networking lets you share printers and other peripherals. It also lets you share files, which means you don't have to worry about getting data out of your old computer?just hook it up to your network. You can even share an Internet connection between computers!    Rate this link
    • Home Networking Tutorial - Home networking is the collection of elements that process, manage, transport, and store information, enabling the connection and integration of multiple computing, control, monitoring, and communication devices in the home. Until recently, the home network has been largely ignored. However, the rapid proliferation of personal computers (PCs) and the Internet in U.S. homes, advancements in telecommunications technology, and progress in the development of smart devices have increasingly emphasized the last 100 feet of any consumer-related network (i.e., the American home). Furthermore, as these growth and advancement trends continue, the need for simple, flexible, and reliable home networks will greatly increase. This tutorial addresses the market drivers, the current and future technologies, and the standards (or lack thereof) relative to home networking and the home-networking environment.    Rate this link
    • How LAN Switches Work - This document covers the general concept of how LAN switches work and the most common features available on a LAN switch. It also covers the differences between bridging, switching, and routing.    Rate this link
    • How Home Networking Works - If you are one these multiple-PC owners, you have probably thought about how great it would be if your computers could talk to each other. This article will look at all of the different methods you can use to create a home network.    Rate this link
    • IEEE 802.2 Logical Link Control (LLC)    Rate this link
    • Introduction to LAN Protocols - This document introduces the various media-access methods, transmission methods, topologies, and devices used in a local-area network (LAN). Topics addressed focus on the methods and devices used in Ethernet/IEEE 802.3, Token Ring/IEEE 802.5, and Fiber Distributed Data Interface (FDDI).    Rate this link
    • LAN Mail Protocols Summary - There are advantages to having a central server receive the mail destined to desktop computers and having the desktop computer collect the mail from the server on demand. There are many protocols designed for this purpose, and this FAQ document takes a look at them.    Rate this link
    • Logical Link Control IEEE 802.2 - The IEEE 802.2 standards for Logical Link Control define a programming interface between that part of the communications software that controls the network interface card (the Media Access Control and Physical Medium Dependent components) and the overlying protocol stack (IP, NetBIOS, NetWare, etc.). The connection between the network interface card and the rest of the communications system is through a structure called a Service Access Point. The SAP differentiates between communications protocols; there's a SAP for NetBIOS, another for SNA, another for NetWare, and so on. A programmer can select Type 2 Logical Link Control in which case the frames are given sequence numbers as they pass through the SAP and the 802.2 Logical Link Control layer at the receiver provides an acknowledgement for received frames. This creates a reliable data transfer mechanism at the Data Link Layer. Type 1 Logical Link Control simply provides the differentiation function, with no sequence and acknowledgement process.    Rate this link
    • Mixed-Media Bridging - Transparent bridges are found predominantly in Ethernet networks, and source-route bridges (SRBs) are found almost exclusively in Token Ring networks. Both transparent bridges and SRBs are popular, so it is reasonable to ask whether a method exists to directly bridge between them. Several solutions haveevolved.    Rate this link
    • Optimized Engineering Technical Compendium - a collection of essays and papers on technical issues of interest to Networking Professionals    Rate this link
    • Rubyan.com - information source on practical tips and tricks for computer network management    Rate this link
    • Service Access Point (SAP) identifiers - These standard SAP numbers are used in the LLC header in all 802.2-compliant protocols like Ethernet to identify which protocol handler should process an incoming frame    Rate this link
    • Source-Route Bridging (SRB) - means to bridge Token Ring LANs    Rate this link
    • State of the LAN: What Ethernet Can't Do - During the past decade, Ethernet's architects - the members of the IEEE 802.3 group - have worked diligently at closing the gap between Ethernet and ATM. They've done an impressive job in an array of areas.    Rate this link
    • The TCP/IP Protocol Family - general introduction    Rate this link
    • Transparent Bridging - background, switching loops, spanning three algorithm    Rate this link
    • VLANs and Broadcast Domains - introduxtory article from Network Magazine    Rate this link
    • VLAN: Virtual Local Area Network and IEEE 802.1Q - Virtual LAN (VLAN) is a group of devices on one or more LANs that are configured so that they can communicate as if they were attached to the same wire, when in fact they are located on a number of different LAN segments. Because VLANs are based on logical instead of physical connections, it is very flexible for user/host management, bandwidth allocation and resource optimization.    Rate this link
    • 802.1Q VLANs for better bandwidth - The 802.1Q specification establishes a standard method for inserting virtual LAN (VLAN) membership information into Ethernet frames. The IEEE's 802.1Q standard was developed to address the problem of how to break large networks into smaller parts so broadcast and multicast traffic wouldn't grab more bandwidth than necessary. The standard also helps provide a higher level of security between segments of internal networks.    Rate this link
    • IEEE 802.1 P,Q - QoS on the MAC level - Purpose of this research paper is to study use of protocols described in IEEE (Institute of Electrical and Electronics Engineers) standards 802.1P and 802.1Q as QoS (Quality of Service) protocols on MAC (Medium Access Control) level. First of all, both standards will be represented, as well as some other related standards as well as some proposals for standards. In this paper, a discussion of future of thise standards will be represented. As well will relations to other QoS, ToS (Type of Service) and CoS (Class of Service) standards. Some practical studies of several implementations of thise protocols will carried out in this paper.    Rate this link

    Ethernet

    Ethernet is alocal-area network (LAN) architecture developed by Xerox Corporation in cooperation with DEC and Intel in 1976. Ethernet uses a bus or star topology and supports data transfer rates of 10 Mbps. The Ethernet specificationserved as the basis for the IEEE 802.3 standard, which specifies thephysical and lower software layers. Ethernet uses the CSMA/CD access methodto handle simultaneous demands. It is one of the most widely implemented LANstandards. Ethernet uses the CSMA/CD access methodto handle simultaneous demands. It is the most widely implemented LAN standards. Ethernet is the most commonly used network protocol - a network language. Ethernet is a type of network cabling and signaling specifications (OSI Model layers 1 [physical] and 2 [data link]).

    Any device connected to a network must have an Ethernet adapter and Ethernet software (usually Ethernet card driver and higher level protocol stacks like TCP/IP). With appropriate network software, any computer can understand and use it. This common protocol and its software enable computers and peripherals to communicate with each other, even if they are using different operating systems. Network software may be provided with a computer or adapter, or it may be purchased separately.

    Ethernet has been a relatively inexpensive, reasonably fast, and very popular LAN technology for several decades. Two individuals at Xerox PARC -- Bob Metcalfe and D.R. Boggs -- developed Ethernet beginning in 1972 and specifications based on this work appeared in IEEE 802.3 in 1980. Ethernet specifications define low-level data transmission protocols and the technology needed to support them. In the OSI model, Ethernet technology exists at the physical and data link layers (layers 1 and 2) .From the time of the first Ethernet standard, the specifications and the rights to build Ethernet technology have been made easily available to anyone. This openness, combined with the ease of use and robustness of the Ethernet system, resulted in a large Ethernet market and is another reason Ethernet is so widely implemented in the computer industry. Most LANs must support a wide variety of computers purchased from different vendors, which requires a high degree of network interoperability of the sort that Ethernet provides.

    Ethernet started as a 10 Mbit/s half-duplex networking technique which used a single coaxial cable as the communication medium. Ethernet has evolved from that to faster and more modern networking technique. Nowadays Ethernet most typically travels over twisted pair wiring or overfiber optic cabling. The typical physical Ethernet network structures are point-to-point links and star network with a HUB in the venter of the star.

    The speed of Ethernet has been updated from 10 Mbit/s to higher speedslike 100 Mbit/s, 1 Gbit/s and 10 Gbit/s. Ethernet support various media. Propagation delays differ between mediums, which affect the maximum possible length of the Ethernet topology running on that medium. In the following table, c refers to the speed of light in a vacuum (300,000 kilometers per second).

            Medium        Propagation Speed
            ------        -----------------
            Thick Coax    .77c (231,000 km/sec)
            Thin Coax     .65c (195,000 km/sec)
            Twisted Pair  .59c (177,000 km/sec)
            Fiber         .66c (198,000 km/sec)
            AUI Cable     .65c (195,000 km/sec)
    

    Ethernet a "broadcast" network. This means that each device connected to the network listens for traffic on the network and then sends its "packets" when the line is quiet. Packets contain sequences of binary information and packet size is usually determined by the application that is running and the type of information that is being transmitted. Packet sizes can range from 64 to 1518 bytes.

    In addition to the data being transmitted, each packet also contains source, destination, and parity (bit error detection) information. The inteframe gap is the amount of time that is specified between frames transmitted from a workstation. The designers of the Ethernet specification arbitrarily chose 96 bit times to occur between frames from a transmitting station (gives some time to perform normal Ethernet housekeeping functions on the network interface card).

    The basic IEEE 802.3 Ethernet MAC Data Frame for 10/100Mbps Ethernet:

    7

    1

    6

    6

    2

    46-1500bytes

    4

    Pre

    SFD

    DA

    SA

    Length Type

    Data unit + pad

    FCS

    • Preamble (PRE)- 7 bytes. The PRE is an alternating pattern of ones and zeros that tells receiving stations that a frame is coming, and that provides a means to synchronize the frame-reception portions of receiving physical layers with the incoming bit stream.
    • Start-of-frame delimiter (SFD)- 1 byte. The SOF is an alternating pattern of ones and zeros, ending with two consecutive 1-bits indicating that the next bit is the left-most bit in the left-most byte of the destination address.
    • Destination address (DA)- 6 bytes. The DA field identifies which station(s) should receive the frame..
    • Source addresses (SA)- 6 bytes. The SA field identifies the sending station.
    • Length/Type- 2 bytes. This field indicates either the number of MAC-client data bytes that are contained in the data field of the frame, or the frame type ID if the frame is assembled using an optional format.
    • Data- Is a sequence of n bytes (46=< n =<1500) of any value. (The total frame minimum is 64bytes.)
    • Frame check sequence (FCS)- 4 bytes. This sequence contains a 32-bit cyclic redundancy check (CRC) value, which is created by the sending MAC and is recalculated by the receiving MAC to check for damaged frames.

    There are two slightly different frame formats used in Ethernet. The Ethernet Version 2 frame format was designed before the IEEE specifications, but is almost identical to the 802.3 frame type. With the Ethernet Version 2 frame type, a two-byte Type field follows the source station's six-byte MAC address. In the 802.3 frame type, this two-byte field after the source address is a length field specifying the number of bytes in the LLC and data fields. If these two bytes are greater than 05DC hex (1500 decimal), the frame is a Version 2 frame. Since all type fields are greater than 1500 decimal (the maximum Ethernet frame size), both frame types can easily coexist on the same network. Some network protocol analyzers call a Version 2 frame an Ethertype frame because of this two-byte Type field.

    This is an Ethernet Version 2 frame:

           +--------------+
           |              | The preamble consists of 62 bits of alternating
           |   Preamble   | ones and zeros that allows the Ethernet card to
           |   7 bytes    | synchronize with the beginning of a frame.
           |              |
           +--------------+ The Start Frame Delimiter is the sequence
           | SFD - 1 byte | 10101011, and indicates the start of a frame.
           +--------------+
           |              | The destination address is a six byte Media Access
           | Destination  | Control (MAC) address, usually burned into the
           |   6 bytes    | ROM of the Ethernet card.
           +--------------+
           |              | The source address is a six byte MAC address, and
           |   Source     | can signify a physical station or a broadcast.
           |   6 bytes    |
           +--------------+
           |     Type     | The Type field (see explanation above).
           |    2 bytes   |
           +--------------+
           |              |  Any higher layer information is placed in the
           |    Data      |  data field, which could contain protocol
           |              |  information or user data.
           ~              ~
           ~              ~
           |  46 to 1500  |
           |    bytes     |
           |              |
           |              |
           +--------------+
           |     FCS      |  The Frame Check Sequence is a cyclic redundancy
           |   4 bytes    |  check used by the sending and receiving stations
           +--------------+  to verify a successful transmission. The FCS is
                             based on the contents of the destination address,
                             source address, type, and data.
    
    
    Frame structure for an 802.3 Ethernet frame that contains the 802.2 LLC information:
           +----------------+
           |                |
           |    Preamble    |
           |    7 bytes     |
           |                |
           +----------------+
           |  SFD - 1 byte  |
           +----------------+
           |                |
           |  Destination   |
           |    6 bytes     |
           +----------------+
           |                |
           |     Source     |
           |    6 bytes     |
           +----------------+
           |  Frame Length  |
           |    2 bytes     |
           +----------------+
           |  DSAP - 1 byte |  The Destination and Source Service Access Point
           +----------------+  fields determine the protocol used for the upper
           |  SSAP - 1 byte |  protocol type of the frame.
           +----------------+
           |Control - 1 byte|  The Control field is used for administration by
           +----------------+  certain protocols.
           |      Data      |
           |                |
           ~                ~
           ~                ~
           |   46 to 1500   |
           |     bytes      |
           |                |
           +----------------+
           |      FCS       |
           |    4 bytes     |
           +----------------+
    

    After the 802.2 frame type was defined, many people felt that a single byte for DSAP and SSAP would not be sufficient to handle the growth of protocols into the future. A single byte DSAP or SSAP can only specify 256 separate protocols, and many of those were predefined from the beginning of the 802.2 specification. To provide future growth, the Sub-Network Access Protocol (SNAP) was created as an extension to the 802.2 specification. To differentiate this protocol from the original 802.2 specification, 802.2 SNAP uses the DSAP and SSAP of 0xAA.

    This is an 802.2 SNAP frame encapsulated in an 802.3 frame:

          +----------------+
           |                |
           |    Preamble    |
           |    7 bytes     |
           |                |
           +----------------+
           |  SFD - 1 byte  |
           +----------------+
           |                |
           |  Destination   |
           |    6 bytes     |
           +----------------+
           |                |
           |     Source     |
           |    6 bytes     |
           +----------------+
           |  Frame Length  |
           |    2 bytes     |
           +----------------+
           |  DSAP - 1 byte |
           +----------------+
           |  SSAP - 1 byte |
           +----------------+
           |Control - 1 byte|
           +----------------+ The Organizationally Unique ID (OUI) is assigned
           | OUI - 3 bytes  | to unique vendors to help differentiate protocols
           |                | from different manufacturers.
           +----------------+
           | Type - 2 bytes | The two-byte protocol type defines a specific
           +----------------+ protocol in the SNAP. This also maintains a
           |                | compatibility with Ethernet v2.
           |      Data      |
           |                |
           ~                ~
           ~                ~
           |   46 to 1500   |
           |     bytes      |
           |                |
           +----------------+
           |      FCS       |
           |    4 bytes     |
           +----------------+
    

    Before the final 802.2 LLC specifications were finalized, Novell implemented IPX/SPX over Ethernet. It originally utilized 802.3 Ethernet without using 802.2 LLC. No other manufacturer uses this frame type. To implement their 'raw' frame type, Novell used the first two bytes of the 802.3 data field as 0xFFFF. Since the DSAP and SSAP values of 0xFF do not exist, it becomes easy to differentiate between the 802.3 and 802.3 'raw' frame types.

    Errors occur when packets do not reach their destination or information is dropped from the data sequence. Common types of errors that may be associated with full network utilization and/or noise disturbances in Ethenet network are:

    • Alignment: Packets do not end on an 8-bit boundary. This is typically caused by noise or broken equipment.
    • Collision: Two devices detect that the network is idle and try to send packets at exactly the same time. Collision errors are common in Ethernet systems and are expected as network utilization increases. Upon receipt of this error type, both devices hold, wait a "randomly" calculated amount of time (to avoid a second collision), and attempt to re-transmit. This is normal operation of (half duplex) Ethernet system.
    • Cyclic Redundancy Check (CRC): Packet size is correct, but the information contained in the frame check sequence (FCS) is corrupt.
    • Fragment: Packet is undersized and contains corrupt FCS.
    • Jabber: Packet is oversized and contains corrupt FCS.
    • Runt/Pygmy: Packets are less than 64 bytes in length.

    Depending upon the severity of the error, the network may ignore packets, re-transmit packets or, the network may halt or 'crash' because the error causes all devices to appear busy.

    Ethernet networking uses collisions as one of the contention access methods. When the network carrier is not active, any station can send information. If two stations attempt to send information at the same time, the signals overlap with each other, creating a collision. Collisions are not errors! Although the term 'collision' may bring to mind a terrible crash, be assured that a collision is a normal part of Ethernet networking. Collisions within the first 512 bits are not errors, they're collisions and entirely normal. Collisions occuring later are due to the failure of a NIC, to detect traffic on the wire.

    When a collision is recognized by a transmitting station, a bit sequence called jam is transmitted. This jam is 32 bits long. Interestingly enough, the actual format of jam is unspecified in the 802.3 specifications. Most manufacturers have used alternating 1s and 0s as jam, which is displayed as 0x5 (0101) or 0xA (1010). In many Fast Ethernet implementations, the jam has been seen as other arbitrary values. A collision is considered late if the jam occurs after 512 bit-times, or 64 bytes. Collisions that occur after the first 64 bytes of a frame may be indicative of a network design problem (the network is so large the jam cannot traverse the entire length in 32 bit-times), or a hardware or Ethernet firmware issue. When collisions do not propagate the network quickly enough, a collision could occur between two stations without the stations aware that the packets collided (frames are simply lost without neworking card knowing it).

    Sometimes you might see term SQE test. The SQE Test is used to test for the collision present circuit between a transceiver and a network interface card (NIC). After data is successfully transmitted, the Ethernet transceiver asserts the SQE signal on the collision presence circuit of the NIC. The NIC sees this test signal as a verification that the transceiver will inform the NIC when a collision occurs. In most modern Ethernet networks, the SQE test is not used or applicable because NICs now have an integrated transceiver (no risk of collosion wire between transceiver and NIC being damaged).

    Ethernet has also evolved from half-duplex bus systems a switched full-duplex networking technique. Ethernet physical connectors provide circuits including the receive (RX), transmit (TX), and collision detection. When half duplex Ethernet is implemented, the TX circuit is always active at the transmitting station. When another station is transmitting, the station's RX circuit is always active. This is referred to as Shared Bandwidth. Standard Ethernet configuration efficiency is typically rated at 50-60 percent of the 10/100Mbps bandwidth. Full duplex Ethernet Switching provides a transmit-circuit connection wired directly to the receiver circuit at the other end of the connection. This two station connected environment creates a collision free situation on the circuit. Recall half duplex Ethernet has to manage the conditions for multiple transmissions on the same physical wire as they cannot occur. Full-duplex operation is possible on networking devices which use twisted pair or fiber wiring and support full-duplex operation. Full duplex Ethernet can operate at up to 100 percent efficiency in both directions. (100Mbps transmit, and 100Mbps receive for example)

    LAN switching is a technique that significantly improves Ethernet network performance without impacting the addressing structure within the network. Switching is defined as the ability to forward packets on the fly through a cross point matrix, a high speed bus, or shared memory arrangement. switch looks at the destination address of each incoming packet, and transmits the packet only on the port on which the destination node is located. Other ports on the switch can transmit or receive different packets at the same time. Besically the idea of swiching has been in Ethernet world for a long time, but the availability of cheap switching devices has made it a mainstream technique (early two-port switches were known as "bridges").

    The current Ethernet is standardized IEEE 802.3 standard.The current edition of IEEE Std. 802.3 is also published as ISO/IEC 8802-3:2000. All approved portions IEEE Std. 802.3 are approved and published at the international level.

    Short history:Ethernet was originally developed by Xerox in the late 1970. A very rarely used 2.94 Mbps version came out of Xerox's Palo Alto Research Center (PARC) in the early 70s. In 1980, Digital Equipment Corporation (DEC), Intel and Xerox published the DIX V1.0 standard which boosted the speed of Ethernet to 10 Mbps while maintaining Ethernet's thick trunk cabling scheme. In 1980, Digital Equipment Corp. (DEC), Intel and Xerox (the origin of the term DIX, as in DEC/Intel/Xerox) began joint promotion of this baseband, CSMA/CD computer communications network over coaxial cabling, and published the "Blue Book Standard" for Ethernet Version 1. This standard was later enhanced and 1982 the enhanced DIX V2.0 specification was released. In 1985 Ethernet II specification based on DIX V2.0 was released. Xerox then relinquished its trademark. At the time of the first DIX standard, the Institute of Electrical and Electronic Engineers (IEEE) was attempting to develop open network standards through the 802 committee. In 1985 the IEEE 802.3 committee published "IEEE 802.3 Carrier Sense Multiple Access with Collision Detection (CSMA/CD) Access Method and Physical Layer Specifications." This technology is called 802.3 CSMA/CD and not Ethernet; however, it is frequently referred to as Ethernet even though the frame definition differs from DIX V2.0. Although 802.3 and DIX frames can coexist on the same cable, interoperability is not assured. Therefore, when discussing "Ethernet," it is necessary to clarify 802.3 frames or DIX V2.0 frames.

    Ethernet history timeline (mostly based on information at http://www.techfest.com/networking/lan/ethernet1.htm):

    • Late 1970's: Xerox developed a new system, Ethernet Version 1, based on ALOHA and capable of providing 3Mbit/s (2.94Mbit/s) throughput. Xerox designed and implemented additional carrier sense and collision detection mechanisms to overcome some of the inherent problems with the original system. This technology was used to connect workstations together over a 1km cable.
    • 1979: Digital Equipment Corporation (DEC), Intel, and Xerox joined for the purpose of standardizing an Ethernet system that any company could use
    • 1980: In September 1980 the three companies released Version 2.0 of the first Ethernet specification called the "Ethernet Blue Book", or "DIX standard" (after the initials of the three companies). It defined the "thick" Ethernet system (10Base5), based on a 10 Mb/s CSMA/CD (Carrier Sense Multiple Access with Collision Detection) protocol. This version of Ethernet is sometimes referred to as DIX Ethernet although the most common name is Ethernet Version 2 or just Ethernet. Version 2 then went on to form the basis of the IEEE 802.3 standard.
    • 1982: The first Ethernet controller boards based on the DIX standard became available. Ethernet Version 2 specification released widely bad became known.
    • 1983: Institute of Electrical and Electronic Engineers (IEEE) released the first IEEE standard for Ethernet technology. It was developed by the 802.3 Working Group of the IEEE 802 Committee. The formal title of the standard was IEEE 802.3 Carrier Sense Multiple Access with Collision Detection (CSMA/CD) Access Method and Physical Layer Specifications. IEEE reworked some portions of the DIX standard, especially in the area of the frame format definition. However the 802.3 standard was defined in a manner that permitted hardware based on the two standards to interoperate on the same Ethernet LAN.
    • 1985: IEEE 802.3a defined a second version of Ethernet called "thin" Ethernet, "cheapernet", or 10Base2. It used a thinner, cheaper coaxial cable that simplified the cabling of the network. Also IEEE 802.3b 10Broad36 standard that defined transmission of 10 Mb/s Ethernet over a "broadband" cable system.
    • 1987: The IEEE 802.3d standard defined the Fiber Optic Inter-Repeater Link (FOIRL) that used two fiber optic cables to extend the maximum distance between 10 Mb/s Ethernet repeaters to 1000 meters. IEEE 802.3e defined a "1 Mb/s" Ethernet standard based on twisted pair wiring (this was never widely used).
    • 1990: Introduction of the IEEE 802.3i 10Base-T standard. It permitted 10 Mb/s Ethernet to operate over simple Category 3 Unshielded Twisted Pair (UTP) cable. This led to a vast expansion in the use of Ethernet.
    • 1993: IEEE 802.3j standard for 10Base-F (FP, FB, & FL) was released which permitted attachment over longer distances (2000 meters) via two fiber optic cables. This standard updated and expanded the earlier FOIRL standard.
    • 1995: IEEE improved the performance of Ethernet technology by a factor of 10 when it released the 100 Mb/s 802.3u 100Base-T standard (commonly known as "Fast Ethernet").
    • 1997: IEEE 802.3x standard became available which defined "full-duplex" Ethernet operation. Full-Duplex Ethernet bypasses the normal CSMA/CD protocol to allow two stations to communicate over a point to point link.
    • 1998: IEEE once again improved the performance of Ethernet technology by a factor of 10 when it released the 1 Gb/s 802.3z 1000Base-X standard. This version of Ethernet is commonly known as "Gigabit Ethernet".
    • 1999: 802.3ab 1000Base-T standard defined 1 Gb/s operation over four pairs of category 5 UTP cabling
    • 2002: IEEE 802.3ae standard introduced 10 Gigabit Ethernet.
    • 2003: IEEE802.3af Draft Standard has been Completed to standardize Power Over Ethernet technology.

    What is an 802.3 network? That's IEEE-ism for Ethernet, but with a few small differences. The physical layer specifications are identical (though DIX Ethernet never specified standards for UTP and Fiber-Optic media) and the MAC sublayer are somewhat different.What is the difference between an Ethernet frame and an IEEE802.3 frame? Why is there a difference? Ethernet was invented at Xerox Palo Alto Research Center and later became an international standard. IEEE handled making it a standard; and their specifications are slightly different from the original Xerox ones. Hence, two different types. 802.3 uses the 802.2 LLC to distinguish among multiple clients, and has a "LENGTH" field where Ethernet has a 2-byte "TYPE" field to distinguish among multiple client protocols. TCP/IP and DECnet (and others) use Ethernet_II framing, which is that which Xerox/PARC originated.

    MAC address is the unique hexadecimal serial number assigned to each Ethernet network device to identify it on the network. With Ethernet devices (as with most other network types), this address is permanently set at the time of manufacturer, though it can usually be changed through software (though this is generally a Very Bad Thing to do). The MAC addresses are exactly 6 bytes in length, and are usually written in hexadecimal as 12:34:56:78:90:AB. Each manufacturer of Ethernet devices applies for a certain range of MAC addresses they can use. The first three bytes of the address determine the manufacturer. RFC-1700 (available via FTP) lists some of the manufacturer-assigned MAC addresses.

    Generation of the data sent to the network and the reception of it is generally done in the combination of software (Ethernet card driver) and hardware (Ethernet card Ethernet controller chip). The Ethernet packet preamble is normally generated by the chipset. Software is responsible for the destination address, source address, type, and data. The chips normally will append the frame check sequence. Ethernet card is highly loaded with traffic when things start slowing down to the point they are no longer acceptable. There is not set percentage point, but you usually start paying attention over 40-50%, or when things slow down .Most typical Ethernet wiring system in use nowadays is one which uses twisted pair wiring. In this system maximum cable distance of 330 feet (100m) between devices. However, the signal can be repeated by either an Ethernet hub or repeater, and this can be done up to 3 times.

    Transparent bridging is a method to connect two similar network segments to each other at the datalink layer. It is done in a way that is transparent to end stations. Transparent bridges are sometimes called learning bridges. When they are turned on and receive data packets from a network segment they:

    • 1) learn the relation between MAC address and segment/port, and
    • 2) forward the packet to all other segments/ports.
    The first step in this process is essential to the "learning" aspect of the bridge. After some time the bridge has learned that a particular MAC address, say MACa, is on a particular segment/port, say PORT1. When it receives a packet destined for the MAC address MACa (from any port not being PORT1) it will no longer forward the packet to all ports (step 2). It knows that MACa is associated with PORT1 and will only forward the packet to PORT1. Originally briges were just normal computers having to (or more) networking cand and running a special software that does the bridging.

    Modern Ethernet switches nowadays typically do this kind of transparent switching. From a functional point of view, switching is exactly the same as bridging. However switches use specially designed hardware to perform the bridging and packet-forwarding functionality. Ethernet switches usually also offer additional capabilities such as virtual LANs (VLANs) and full duplex connectivity.

    Spanning tree is a protocol defined in IEEE 802.1D to prevent bridges from creating network loops. Using the spanning tree protocol, bridges communicate to each other and disable certain ports/segments to prevent looping of packets. Many implementations of the spanning tree protocol are configured so an alternate path is available to network traffic, should the original path become disabled.

    To further confuse issues, standard Ethernet sometimes in marketing speaks means an attached protocol- mainly TCP/IP. Ethernet only defines the data link and physical layers of the Open Systems Interconnect (OSI) Reference Model whereas TCP/IP defines the transport and network layers respectively of the same model.

      Ethernet standards

      • Get IEEE 802 - The Get IEEE 802 program makes IEEE 802? standards (like IEEE 802.3 Ethernet) available at no charge in PDF format. This pilot program grants public access to view and download current individual IEEE Local and Metropolitan Area Network standards at no charge. New IEEE 802? standards are included in the program after they have been published in PDF for six months.    Rate this link

      Ethernet switching

      Ethernet switching provides the speed, performance, bandwidth, and flexibility required for today's enterprise networks. Switches are a fundamental part of most networks. They make it possible for several users to send information over a network at the same time without slowing each other down. Switches allow different nodes (a network connection point, typically a computer) of a network to communicate directly with each other in a smooth and efficient manner.

      There are a lot of different types of switches. Switches that provide a separate connection for each node in a company's internal network are called LAN switches. Essentially, a LAN switch creates a series of instant networks that contain only the two devices communicating with each other at that particular moment. Ethernet switching products are originally designed to deliver Layer 2 connectivity, althrough some products nowadays offer also Layer 3-7 content-aware intelligence. Layer 2 switching provides the dedicated bandwidth and network segmentation critical for directly connecting users to the network. Layer 3 provides for switching and routing, maximizing speed, bandwidth, and flexibility in the network core or aggregation points.

      There are three main techniques for Ethernet switching:

      • Store and Forward: Ethernet switch receives the full frame to it's memory and then decides what to do with it.
      • Cut Through: Switch makes the decision on re-transmission when it has received the destination MAC address (this resides on the header of the frame)
      • Fragement Free (Modified Cut Through): Switch makes the decision on re-transmission after it has received the first 64 bytes of the frame.
      Switches work typically in Store and Forward or Cut Through modes. Both methods have their good sides. Some switches cna operate at the both modes depending on the situation. When Ethernet switches handle ports that operate at different speeds, some form of flow control would be useful. The standard approach in Ethernet devices is to drop the packets that come in too fast. For this reason 802.3x standard added flow constolling to Ethernet standard: the device which is overloaded can send PAUSE commands to the transmitting equipment. This standardized flo control works only on full-duplex switched Ethernet connections. In case some form of floc controlling is needed in connections that use original CSMA/CD protocol, the typical approach here is to use non-standard "Back Pressure" protocol: switch sends indications of collision to the traffic source.

      Ethernet switches are available on large variery of versions from the simple switches without any fancy features to very feature rich managed switches. Some typical switches:

      • Managed switches are the switches that offer you manageblity through some suitable control interface. This managebility usually includes configuring the operation mode of different ports (speed, full/half duplex operation, auto detection on/off) and how traffic is switched (security features, virtual LANs, possibly bandwidth control). The namagement of managed switches is typically done using methods like telnet connection, through serial port on the device, HTTP user interface and/or SNMP protocol (which of those different devices support vary). You can usually get informtion on state of the device through the management interface.
      • Unmanaged consumer level switches are very simple devices. There is no http interface, no GUI, no telnet, no configurability at all other than what can be done with switches on the front panel. The consumer level switches typically just switch the traffic from port to another without any extra controls and auto-detect the operation mode of different ports.

      To get functionality than what cheap basic switches can give, you will need a managed switch, which is far, far more costly than simple basic consumer switches. The price range for cheap consumer switches stat from just few tens of dollars/euros (just few dollars/euros per port). The price range for managed switches is typically from few hundred dollars to few thousand dollars depending on the number of ports and other features.

      Ethernet switching adds network security by making sure that the packets go only to those destinations that need those packets. This adds security greatly, although the security is not foolproof (for example broadcast messages get everywhere and in some cases packets can get to other destinations also than where they are ment to go). Network monitoring/snooping used to be so easy on networks before switches.

      Network traffic analyzing (sniffing) is harder with switched networks, because sniffer can analyze packets it can't see. Some switches can be configured to monitor a port, which can help in some cases. It's relatively common on managed switches to offer a port "mirroring" feature, which copies port traffic to a different location. Nortel calls it mirroring; Cisco calls it "SPAN" if the data is sent to a local port, "RSPAN" if the traffic is sent remotely. The selection criteria for this copying vary greatly between manufacturers and models; for some it copies everything always; others allow you to be selective with criteria such as source port, source IP, destination port, destination IP, protocol, or VLAN tag. In some switches, the destination port the traffic is being copied to is isolated from everything else and will only transmit the copied data, while on some opther switches the destination port can still be used for regular traffic. Different switches also differ on two other important features: whether VLAN tags get stripped off; and whether the original source MAC address of the packet is preserved or if the original source MAC is replaced with the MAC of the egress port of the switch. In some applications you can use a little 'pocket' hub that you can drag with you and then route the segments through the hub, and place the snooping device on the hub (when usign this please note that some cheapo stuff with a "hub" badge coul?d be really a switch). You can try to flood the switch with (faked) arp-packets causing the switch to act like an hub, but this will definetly influence any attempt to do some troubleshooting.

      • Bridging and Switching Basics - Bridges and switches are data communications devices that operate principally at Layer 2 of the OSI reference model. As such, they are widely referred to as data link layer devices.    Rate this link
      • How LAN Switches Work - This document covers the general concept of how LAN switches work and the most common features available on a LAN switch. It also covers the differences between bridging, switching, and routing.    Rate this link

      10Base-T information

      The "T" in 10BASE-T stands for "twisted" in reference to the twisted-pair wire used for this variety of Ethernet. This is nowadays the most commonly used Ethernet variety (thanks to the popularity of structured cabling systems which are based on twisted pair wiring). The specifications for the twisted-pair media system were published in 1990. This system has since become the most widely used medium for connections to the desktop. The 10BASE-T system was the first popular twisted-pair Ethernet system.

      The invention of 10BASE-T in the early 1990s led to the widespread adoption of Ethernet for desktop computers.10BASE-T system is designed to work with unshielded twisted pair wiring with impedance of 100 ohms + or - 15%. The maximun link length is speified to be 100 meters when using data grade cable (category 3 or better). The 10BASE-T system is designed to support the transmission of 10 Mbps Ethernet signals over "voice-grade" Category 3 twisted-pair cables. However, the vast majority of twisted-pair cabling systems in use today are based on Category 5 twisted-pair cables. Category 5 cables have higher quality signal carrying characteristics and work very well with the 10BASE-T system.

      The 10BASE-T system operates over two pairs of wires, one pair used for receive data signals and the other pair used for transmit data signals. The two wires in each pair must be twisted together for the entire length of the segment, a standard technique used to improve the signal carrying characteristics of a wire pair. 10BaseT Ethernet is a baseband system that uses a Manchester encoding of high and low voltages to place bits on a wire pair. It a Manchester encoding each bit time contains a transition in the middle: a transition from low to high represents a 0 bit and a transition from high to low represents a 1 bit. With repeated bits of the same value, a transition is also needed at the edge of the bit time. To achieve 10 Mbps, one needs a capacity for 20 million transitions per second. Output signal level of a typical 10Base-T ethernet device is 2.2V to 2.8V (leads to around 2Vpp on each of the wires in the pair). The input signal level in the receiver end should be from 350mV (minimum signal level) to 2.8V (maximum ethernet card signal level). Most of the energy sent to the line is in 5-15 MHz frequency range. The 10Base-T Ethernet is designed for 100 ohm cable impedance (85-115 ohms allowed) cabbling. The standard Ethernet connector is 8-pin modular connector also known as RJ-45.

      Ethernet 10BaseT pinout:
      Pin #Signal NameFunctionEIA/TIA T568B color
      1TD+Transmit Data white-orange
      2TD-Transmit Data orange
      3RD+Receive Data white-green
      4NCNo Connection blue
      5NCNo Connection white-blue
      6RD-Receive Data green
      7NCNo Connection white-brown
      8NCNo Connection brown
      NOTE: T568A wiring changes the position of green and orange wire pairs.

      Signaling in 10BASE-T Ethernet is performed using Manchester phase encoding. In a "phase encoded" signal, the logic state (0 or 1) is indicated by the phase of the carrier signal, rather than by a fixed voltage level as in standard logic circuits. In "Manchester phase encoding" the data bit rate is the same as the base frequency of the master clock oscillator (10 MHz for standard 10BASE-T Ethernet). A data bit 0 in the level encoded signal is represented in the phase encoded signal by a full cycle of the original clock, while a data bit 1 is represented by a full cycle of the inverted clock. This encoding technique has the advantage that regardless of the data being transmitted, the encoded data have regular transitions, with a maximum time of one clock period between transitions. When this manchester code is sent ot the cable there are few signal tricks used to minimize the EMI problems and guarantee that the signal goes nicely through the cable:

      • Pre-distorion which equalizes the signal so that around 50 meters of cable (acts as lowe pass filter) makes signal back to original, so the signal distortions are not too large in any part of cable (either some pre-distortion or some cable filtering)
      • Signal is send differentially to cable and is well balanced (good noise cancelling and rediced emissions)
      • The signal is sent to the cable through a balancing transformer (safety and good common mode rejection characteristics)
      • Common mode filters are used in the input and output wires (reduces conducted emissions)

      For 10Base-T Ethernet the fundamental frequency will be between 5 MHz (alternating ones/zeros) and 10 MHz (all ones/all zeros). The energy spectrum of a packetized Ethernet signal using Manchester encoding at 10 Mb/s is concentrated under 30 MHz, with signal energy down to (but not including) DC. For 10baseT, though, the energy is pretty low at lower frequencies, becaus signal still have to get through the transformers. The maxumum energy is aroud frequencies between around 5-20 MHz (frequencies 5 MHz and 10 MHz being the strongest components). For example for a 10 Mbps Ethernet LAN, the preamble sequence encodes to a 5 MHz square wave.

      Generally 10Base-T ethernet cards use RJ-45 connector to do the connection to the twisted pair wiring (some early card coud have had options for other connectors also). The transmitted data from the Ethernet card leaves at the wire pair which connects to RJ-45 connector pins 1 (Tx+) and 2 (Tx-). The received data comes to the Ethernet card through a twisted pair which connects to the RJ-45 connector pins 3 (Rx-) and 6 (Rc-). 10Base-T ethernet card can be directly wired to s structured cabling system wired according EIA/TIA 568A/568B and/or AT&T 258A. Because standard structured cabling systems use four wire pairs, and Ethernet uses only two of the, there are two wire pairs left to be used for other applications if needed. Those other applications could be for example putting another Ethernet connection to same cable or putting telephone signals to them (those work but are not recommended practices). In some applications those extra wires could be used to supply power to some Ethernet devices that take power through network wiring (modern Power over Ethernet standard devices and some older proprietary systems to power WLAN base stations and IP phones thropugh network cabling).

      10BASE-T is point-to-point technique, which means that one wire can only connect two devices (two computer directly or computer to a HUB). This wire and devices on the end of it form one Ethernet segment. Multiple twisted-pair segments communicate by way of a multiport hub.

      10BASE-T MAUs continually monitor the receive data path for activity as a means of checking that the link is working correctly. When the network is idle, the MAUs (network cards or transceivers) also send a link test signal to one another to verify link integrity. Vendors can optionally provide a link light on the MAU; if the link lights on both MAUs are lit when you connect a segment, then you have an indication that the segment is wired correctly. 10BaseT NICs uses a single normal link pulse (NLP) to perform a link integrity test. Typically an indicator LED on the NIC showes the status of the link. NLP pulses are typically generated every 16 ms when the transmitter is idle. Link LEDs are very useful intetwork faulfinding, but they are not fool-proof. Please note that link LEDs do not always guarantee that the wire link work for real traffic. Since the link test pulse operates more slowly than actual Ethernet signals, the link lights are not always a guarantee that Ethernet signals will work over the segment

      You can connect 10BaseT interface of two devices directly together, without using a hub. To do this you need a "crossover cable" that crosses the data transmit and receive pairs.

      Thick Coaxial Ethernet

      The thick coaxial media (10BASE-5) system was the first media system specified in the original Ethernet standard of 1980. Thick coaxial segments are still sometimes installed as a backbone segment for interconnecting Ethernet hubs, since thick coaxial media provides a low-cost cable with good electrical shielding that can carry signals relatively long distances between hubs. Thick coaxial cable is limited to carrying 10-Mbps signals only. Thick coaxial segments can only be connected in the bus cable form of physical topology. The maximum lenght of the cable segment is 500 meters. The cable impedance is 50 ohms. Coaxial cable for use in 10BASE-5 is double-shielded 0.4 inch diameter RG8 coaxial cable (about the size of a garden hose). The cable is not flexible, and difficult to work with. The cable has a characteristic impedance of 50 ohms.

      Thick Ethernet coaxial cable bus must be terminated. The standard termination is 50 +/-2 ohms. The end connector on the RG-8 cable is an "N" type connector. The cable is externally terminated with a resistor inside an N connector. Proper termination is essential for the proper operation of the network, because missing or wring termination causes signal reflections and causes that the signal collision detection does not work properly. The standard notes that the thick coax segment should be grounded at one point for electrical safety reasons. There must only be only one grounding point, to avoid disrupting the Ethernet signals carried by the cable. All other metal parts on the cable should be insulated or carefully routed and fastened in place with plastic cable ties to avoid accidentally touching an electrical ground.

      Ethernet 10Base5 Characteristics in coaxial cable:

      • Output voltage: Voh = -0.225V and Vol = - 1.828V
      • Output current AC: +-16 mA nominal (14-19 mA allowed range)
      • Output current DC: +4.5 mA (4-5 mA allowed range)
      • Output impecance: Grater than 50 kohms
      • Circuit type: ECL
      • Level transistion time: 25 ns +- 5 ns @ 10-90% transistion
      • Encoding format: Manchester phase encoding
      • Transmit/recevie frequency: 10 Mbps +-0.01%
      • Topology: Branching bus
      • Medium: Shielded coaxial cable (50 ohm impedance, for example RG58 cable)
      • Access control: Carrier Sense, Multiple Access with Collision

      For 10 Mb/s Ethernet (using Manchester encoding) the fundamental frequency will be between 5 MHz (alternating ones/zeros) and 10 MHz (all ones/all zeros). The energy spectrum of a packetized Ethernet signal using Manchester encoding at 10 Mb/s is concentrated under 30 MHz. The energy goes down to DC for coaxial ethernet.

      An Ethernet interface is attached to a thick Ethernet segment with an external MAU. The MAU provides an electrical connection to the thick Ethernet coax and transfers signals between the Ethernet interface and the network segment. The MAU physically and electrically attaches to the coaxial cable by a cable tap. The cable is pierced, and a connection is made (by a screw) to the center conductor. The specifications state that there may be a maximum of 100 MAUs attached to a segment, and that each MAU connection to the thick coax be placed on any one of the black bands marked on the coaxial cable. An AUI cable can be used to provide the connection between an external MAU and the Ethernet interface. The MAU is equipped with a male 15-pin connector with locking posts, and the Ethernet interface (DTE) is equipped with a female 15-pin connector that is typically provided with a sliding latch.

      Characteritics of MAU interace:

      • Signal type: Digital
      • Output voltage: Voh = +700 mV and Vol = - 700 mV
      • Signal type: AC signal
      • Circuit type: Balanced, differential ECL
      • Encoding format: Manchester phase encoding
      • Transmit/receive frequency: 10 Mbps +-0.01%
      • Topology: Chained (point-to-point from card to MAU)
      • Medium: Shielded multiconductor cable (78 Ohm balanced shielded twisted pair connections), 15 pin shielded connector
      • Access control: Carrier Sense, Multiple Access with Collision Detect (CSMA/CD)
      • Power from Ethernet card to MAU: +12V

      The 10BASE5 Ethernet interface connector passes three pairs of transformer isolated differential signals for signal transmission, reception, and collision detection. The differential signals TX+/-, RX+/-, and COL+/- at the Serial Interface Adapter (SIA) chip may have risetimes approaching 2 nanoseconds. The interface also supplies 12 volts DC to enable the use of externally powered media access units (MAUs).

      Pinout of 15-pin D connector used in MAU connection:

      • 1 Shield
      • 2 Collision Detect A
      • 3 Transmit Data A
      • 4 (Receive Data Shield)
      • 5 Receive Data A
      • 6 Vc
      • 7 (Control Out A)
      • 8 (Control Out Shield)
      • 9 Collision Detect B
      • 10 Transmit Data B
      • 11 (Transmit Data Shield)
      • 12 Receive Data B
      • 13 V+
      • 14 (V Shield)
      • 15 (Control Out B)

      The standard AUI cable is relatively thick (approx. 1cm or 0.4 inch diameter), and may be up to 50 meters (164 feet) long. The maximum allowable length between a station and a MAU for thinner "office grade" AUI cables is 12.5 meters (41 feet). A typical MAU cable consists of four shielded twisted pair wires that carry the power, transmit data, receive data and collision detect information.

      10Base5 spec says the coax SHALL be grounded at one and only one point. Grounding your coax is generally a good idea; it allows static electricity to bleed off and, supposedly, makes for a safer installation. Further, many local electrical codes will require your network cabling to be grounded at some point. When you ground your Ethernet cable, make sure you do so only at one point. Multiple grounds on an Ethernet segment will not only cause network errors, but also risk damage to equipment and injury to people. If you have a repeater on one end of the segment, this will usually automatically ground that end of the segment (you may want to check the repeater documentation and configuration to assure this is the case?most repeaters can be set-up to NOT ground). If you don?t have a repeater, you can get terminating resistors with ground straps attached.

      Thin coaxial Ethernet

      The thin coaxial Ethernet system uses a flexible coaxial cable that makes it possible to connect the coaxial cable directly to the Ethernet interface in the computer. The cable used is typically RG-58 coaxial cable or similar. The cable is teriminated with BNC connecors. Thin coaxial cable is limited to carrying 10-Mbps signals only.

      Thin coaxial segments can only be connected in the bus cable form of physical topology. The maximum lenght of the cable segment is 185 meters. The cable impedance is 50 ohms. Up to 30 MAUs are allowed on each thin Ethernet segment. The standard requires that multiple segments of thin coaxial cable be linked with repeaters. Each end of a complete thin Ethernet segment must be equipped with a 50 ohm terminating resistance. It is essential that exactly two 50 ohm terminators be installed or enabled on a given segment, or the collision detection mechanism in the MAUs attached to the segment will not function correctly.

      The cable is connected to the Ethernet cable using BNC Tee that is connected directly to the female BNC on the interface. The standard notes that the length of the "stub" connection from the BNC MDI on the interface to the coaxial cable should be no longer than four centimeters (1.57 inches), to prevent the occurrence of signal reflections which can cause frame errors.

      Ethernet 10Base2 Characteristics:

      • Output voltage: Voh = -0.225V and Vol = - 1.828V
      • Output current AC: +-16 mA nominal (14-19 mA allowed range)
      • Output current DC: +4.5 mA (4-5 mA allowed range)
      • Output impecance: Grater than 50 kohms
      • Circuit type: ECL
      • Level transistion time: 25 ns +- 5 ns @ 10-90% transistion
      • Encoding format: Manchester phase encoding
      • Transmit/recevie frequency: 10 Mbps +-0.01%
      • Topology: Branching bus
      • Medium: Shielded coaxial cable (50 ohm impedance, for example RG58 cable)
      • Access control: Carrier Sense, Multiple Access with Collision

      For 10 Mb/s Ethernet (using Manchester encoding) the fundamental frequency will be between 5 MHz (alternating ones/zeros) and 10 MHz (all ones/all zeros). The energy spectrum of a packetized Ethernet signal using Manchester encoding at 10 Mb/s is concentrated under 30 MHz. The energy goes down to DC for coaxial ethernet.

      The Ethernet standard notes that you may provide a thin coaxial segment with a grounding point for electrical safety. To avoid disrupting the Ethernet signals carried by the cable, there must only be one grounding point. All other metal parts on the cable should be insulated or carefully routed and fastened in place with plastic cable ties to avoid accidentally touching an electrical ground. ?

      The flexibility and low cost of the thin coaxial system has made it popular for networking clusters of workstations in an open lab setting, although twisted pair wiring s catching there also quicly also.

      When making the wiring for 10Base-2 network, be sure your cable is the correct impedance. The correct cable impedance is 50 ohms. The right coaxial cable type to use is RG-58, although some other 50 ohm coaxial cable types can be also used (as long as they have similar characteristics as RG-58 and you have BNC connector that you can terminate that cable). Make sure that the connectors and terminators match the impedance of the cable. If you use wrong impedance cable for ethernet wiring, you'll get subtle and intermittant data errors that are tough to track down. When building the cables, buy and use a good crimping tool. Do not use those twist-on connectors. They loosen and bring more intermittant problems.

      Use "T" connectors to connect the PC to the cable segment. Connect the "T" directly to the PC. You aren't allowed to run a cable from the "T" to the PC.Remeber that you must have at least 4.5 feet (1.5 meters) of cable between PCs. One (and only one) end of the cable should be grounded. It's generally easy enough to do this by running a ground line from the terminator to some reliable grounding point (for example mains power ground). If you ground both ends or don't ground either end the result will be intermittant problems. The thin Ethernet standard is designed for 50 ohm coaxial cable and this is the cable type which should ne used. Be careful to make reliable connections all the way through the network bus, because one bad connection on the route will make the whole network bus segment not working.

      In some applications in Ethernet it is desirable to connect the DTE to a non-standard impedance coaxial cable. The 2 most commonly used non-standard cables have 75 Ohm and 93 Ohm impedances. The major difference between non standard cables comes down to segment length. This is due to the increase in cable resistance. Because the 10Base2 Ethernet transceiver is a current driver, the parameters primarily effected by the change in cable impedance are the transmitter and collision voltage detection levels. Some very old Ethernet cards (from 1980-1990) have also supported non-standard 75 ohm coaxial cable as a wiring option. This 75 ohm cable is very rarely used.

      IEEE 802.3 specifies maximum cable length to be 500 meters for 10BASE-5 and 185 meters for 10BASE-2. Extension of this maximum cable length to 1000m and 300m respectively is made possible by means of a technique referred to as, "Transmit Mode Collision Detect." In this scheme it is key that the transmitting node be assured of detecting its own collision and not those of all the stations on the cable. When used, this method allows longer cable segments. Those extra-long cable segment options are not normally used in Ethernet systems.

      The 10Base2 spec says the coax MAY be grounded at one and only one point. Grounding your coax is generally a good idea; it allows static electricity to bleed off and, supposedly, makes for a safer installation. Further, many local electrical codes will require your network cabling to be grounded at some point. The grounding of 10Base2 network is not absolute necessary for the wiring to work. You should absolutely install cabling according to your electrical codes. When you do ground your cable, make sure you do so only at one point. Multiple grounds on an Ethernet segment will not only cause network errors, but also risk damage to equipment and injury to people. In thin Ethernet cabling multiple groungs can happen if the "T" piece and the connectors attached to it touch the PC case or metallic shield of some other cable on the back of the PC. If there is considerable risk of getting multiple ground connections in this way, it is a good idea to use insulated BNC connectors and T pieces when you build your network. If you have a repeater on one end of the segment, this will usually automatically ground that end of the segment (you may want to check the repeater documentation and configuration to assure this is the case?most repeaters can be set-up to NOT ground). If you don?t have a repeater, you can get terminating resistors with ground straps attached.

      10Broad36

      10BROAD36 is a seldom used Ethernet specification which uses a physical medium similar to cable television, with CATV-type cables, taps, connectors, and amplifiers. 10BROAD36 is the only 802.3 broadband media specisfication. It uses 75 ohm CATV coax as the medium. 10Broad36, which is part of the IEEE 802.3 specification, has a distance limit of 2.24 miles (3600 meters) per networkt. Single 10Broad36 segments can be as long as 1800 meters. All 10Broad36 networks are terminated by a "head end" device. Broadband cable systems like CATV support transmission of multiple services over a single cable by dividing the bandwidth into separate frequencies, with each frequency assigned to a different service. This capability can allow 10Broad36 share a single cable with other services such as video. Broadband is by nature analog, so analog encoding must be used.10Broax36 uses PSK modulated radio frequency (RF). The transmission rate is 10 Mb/s. The broadband MAU uses a data band 14 MHz wide and an adjacent collision enforcement band 4 MHzBroadband transmission differs from baseband transmission in the direction of signal flow. The signal moves in only one direction along the cable. In order for signals to reach all the devices in the network, there must be two paths for data flow. This may be accomplished through either a "single cable" or "dual cable" configuration. On a dual-cable system the transmitand receive carrier frequencies are identical and the MAU connects to the medium via two taps,one on the receive cable and the other on the transmit cable. CATV-type broadband cable installation is typically a single bidirectional cable with bandsplitamplifiers and filters. In single cable system te physical tap is a passive directional device such that the MAU transmission is directed toward the headendlocation (reverse direction). On a single-cable system the transmission from the MAU is at a carrier frequencyf1. A frequency translator (or remodulator) located at the headend up-converts to a carrier frequencyf2, which is sent in the forward direction to the taps (receiver inputs). A single cable midsplit con.guration with a frequency offset of 156.25 MHz or 192.25 MHz between forward and reverse channels is recommended.The collision detection in 10Broad36 is quite special: A transmitting MAU logically compares the beginning of the received data with the data transmitted. Anydifference between them, which may be due to errors caused by colliding transmissions, or reception of anearlier transmission from another MAU, or a bit error on the channel, is interpreted as a collision.When a collision is recognized, the MAU stops transmission in the data band and begins transmission of anRF collision enforcement (CE) signal in a separate CE band adjacent to the data band. The CE signal isdetected by all MAUs and informs them that a collision has occurred. All MAUs signal to their attachedMedium Access Controllers (MACs) the presence of the collision. The transmitting MACs then begin the collision handling process. Collision enforcement is necessary because RF data signals from different MAUs on the broadband cable system may be received at different power levels.When introduced, 10Broad36 offered the advantage of supporting much longer segment lengths than 10Base5 and 10Base2. But this advantage was diminished with introduction of the fiber based FOIRL and 10Base-F standards. 10Broad36 is not capable of supporting the full-duplex mode of operation.

      Metropolitan Access Ethernet

      Ethernet technology is designed originally or LAN, but it's usage has expended to campus networks and metropolitan area networks. Outside of enterprise networks, Ethernet is beginning to catch on as a means of Internet access and for connecting metropolitan LANs. But Ethernet has stalled there because it lacks some of the features, particularly in quality-of-service, needed to provide private-line services. Several technologies are coming together to make Ethernet compelling for metro access. Sonet and Ethernet, two of the networking industry's most popular standards, are uniting in a way that might be inevitable given recession-era carrier trends. The idea of using Ethernet throughout the network - replacing Sonet entirely - appears to have vanished, but equipment vendors are still keen on selling the technology as a means of creating services at the network edge. Many metropolitan networks nowadays are based on Sonet/SHD technologues. Carriers built the metro using mature Sonet technologies, which, while optimal for voice or other jitter- and delay-sensitive applications, lack fast circuit-provisioning capabilities, scalability and bandwidth efficiency. This makes the MAN inefficient for the cost-effective transport of data. To leverage Ethernet in the metro it is often necessary to understand the existing Sonet/SDH infrastructure and how it can be adopted to take Ethernet traffic efficiently. With recent innovations in Sonet/SDH and metro Ethernet, the perfect storm of technologies has been brewed to offer Ethernet essentially over any distance. Switched Ethernet over Sonet is emerging as a viable way to migrate to packet-switched nets while preserving the current infrastructure. The advent of generic framing procedure (GFP) and related standards promise the ability to merge the worlds of Ethernet and Sonet more efficiently. One of the most interesting technologies is the combination of virtual concatenation (VCAT) and generic framing procedure (GFP) in the Sonet/SDH segment, along with the enhancements to Ethernet that make it "carrier-worthy." VCAT comes in two varieties: High-order VCAT and Low-order VCAT. For high-order VCAT, two STS-1 data paths could be grouped to yield a 100-Mbit point-to-point Ethernet network that spans any distance. For low-order, seven VT1.5 tributaries could be grouped to create a cost-effective 10-Mbit Ethernet point-to-point network. VCAT can be incorporated into an existing Sonet/SDH network by adding technology at the end points. Another technology crucial to offering metro Ethernet services is the generic framing procedure. GFP is standardized by the ITU as G.7041 and describes the encapsulation and data-rate adaptation techniques for transporting various protocols over Sonet/SDH networks. GFP offers two categories of service: framed and transparent. Framed GFP packages a complete Ethernet (or other) frame into a GFP header. It is important to realize that the frame is carried in its entirety; thus, to the end user, it appears that the Ethernet network is expanded and can be managed like a large enterprise network. Transparent GFP creates a data pipe that moves 8B/10B encoded data from end to end in a streaming fashion. Streams of 8B/10B traffic are encoded into 64B/65B superblocks for transport over the Sonet/SDH network. Rate adaptation is achieved by inserting and removing idle characters. Today, framed GFP is standardized for Gigabit Ethernet. Other protocols, including Fibre Channel and lower-speed Ethernets, will be standardized in the future. Today it is possible to transport Gigabit Ethernet, 1- and 2-Gbit Fibre Channel, Ficon, Escon/Sbcon and DVB-ASI over transparent GFP. In a typical scenario a service provider offers 10/100 Ethernet virtual private network (VPN) services by aggregating the 10/100 Ethernet traffic over a Gigabit Ethernet connection. Aggregated Ethernet flows can be distinguished by inspecting parts of the Ethernet frame. By examining the outer-most virtual LAN (VLAN) tag, multiprotocol label switching (MPLS) label, Internet Protocol (IP) type-of-service byte, DiffServ code point or Ether-net source address (or combinations of these), we can use a simple table lookup to determine the Sonet/SDH VCAT group into which the flow should be encapsulated. If the incoming traffic already has labels that can collide with the ones used by the operator, there is sometimes needs to switch or stack labels. Ethernet VLANs can be stacked using the relatively new "Q in Q" label-stacking scheme, so named for VLAN standard IEEE 802.1Q. MPLS labels have had stacking capabilities from the start. One challenge that will be encountered in moving Ethernet to the metro is the need for lossless flow control. Ethernet comes equipped with the ability to send pause frames once a watermark is tripped. To preserve a loss-free environment in metro applications, there needs to be enough buffering to hold up to three jumbo frames (9,600 bytes) per interface to accommodate a span of 10 km. Ethernet has become so economical that using a link to less than its throughput capacity has a very minor impact on the total cost of the solution. The bulk of the cost is still in the Sonet/SDH part of the network. So the operatotr might offer service like 10-Mbit Ethernet connection that is policed down to 3 Mbits/s. Provisioning is always provided by, at a minimum, specification of a committed information rate (CIR). Traffic that complies with its CIR is always delivered. Provisioning can also be offered using a burst information rate (BIR). Here, traffic that exceeds the CIR parameter but is less than the BIR parameter is delivered on a "best effort" basis.

      Power over Ethernet

      Power over Ethernet (PoE) is a technology for wired Ethernet LANs (local area networks) that allows the electrical current, necessary for the operation of ach device, to be carried by the data cables rather than by power cords. The idea for supplying power and data shared lines are not new - they were also shared in the days of the telegraph and are shared in normal telephone lines (PSTN) for very many decades. Sharing the same wires on the LAN environment is much newer technology. Traditionally the LAN cabling has provided thew data connectivity and all the computer devices connected to it have received the power from nearby mains wall outlet. The history of the idea of using LAN cabling to supply also power to devices connected to it seems to be around 15 years or so (Patent US4733389 shows the PoE concept for Ethernet).

      PoE has gained the lage interrest only on last several years with introduction of a large variaety of WLAN products. There are now components that allow power to be sent over Ethernet data cables. Those allow some low power devices like IP phones and wireless access points to be powered through LAN cabling.The use of same wires for power and data minimizes the number of wires that must be strung in order to install the network. Power over Ethernet allows for example next generation IP telephones, wireless LAN and Bluetooth access points to share power and data over the same cable so that they do not require additional AC wiring or external power transformers.

      When power-over-Ethernet technology becomes widespread, corporations and businesses will see at least one set of wiring hassles disappear or reduced considerably. The real beauty of the standard is that Power over Ethernet is completely compatible with existing Ethernet switches and networked devices. Because the Power Sourcing Equipment (PSE) tests whether a networked device is PoE-capable, power is never transmitted unless a Powered Device is at other end of the cable. It also continues to monitor the channel. If the Powered Device does not draw a minimum current, perhaps because it has been unplugged or physically turned off, the PSE shuts down the power to that port. Optionally, the standard permits Powered Devices to signal to the PSEs exactly how much power they need. For PoE to work, the electrical current must go into the data cable at the power-supply end, and come out at the device end, in such a way that the current is kept separate from the data signal so neither interferes with the other.

      The current enters the cable by means of a component called an injector or PoE power feeding HUB. If the device at the other end of the cable is PoE compatible, then that device will function properly without modification. If the device is not PoE compatible, then a component called a picker or tap must be installed to remove the current from the cable. This 'picked-off' current is routed to the power jack. To minimize the possibility of damage to equipment in the event of a malfunction, the more sophisticated PoE systems employ fault protection. This feature shuts off the power supply if excessive current or a short circuit is detected. Many modern intelligent PoE devices are also designed so that they will supply the power to the line only if there is a PoE capable device connected to line (there is a low-power OpE device detecting scheme). In this way the devices that are not PoE capable will not get the power and there is no changes of any equipment damage. The IEEE 802.3af standard committee has selected a resistor method of detection.

      There has been several prorietary technologues for PoE in use by many equipment manufacturers, but stadardization on this field is coming fast. The IEEE 802.3af specification defines how to transport the pover through Ethernet wiring. The idea on Power-over-Ethernet (PoE) is to feed the power using the same set of wires that carry the data. The power can be carried over the normally unused wires in 4-pair on CAT5 wiring (normally only two pairs are in use, two pairs not used in 10/100 Mbit/s Ethernet). Power can be also fed through the same wires as the data. In-Band-Powering is when the Power over Ethernet power is supplied on the same wires as the Data Pairs of the Ethernet cable (on pins 1&2, 3&6 on 10Base-T and 100Base-TX systems, Gigabit Ethernet uses all four pairs). Out-of-Band-Powering is when the Power over Ethernet power is supplied on the Spare wires of the Ethernet cable (on pins 4&5, 7&8 that are spare pairs on 10Base-T and 100Base-TX systems).

      Power-Over-LAN or Power-Via-MDI (Media Dependent Interface) or Power-over-Ethernet (PoE) is a network infrastructure to deliver 48V power over the existing network Category 5 (CAT-5) data cable. Power-Over-LAN provides signals as well as power to the connected power devices (PDs) such as IP Phones and Wireless Access Nodes, eliminating the need for local power sources. The power that is fed to the devices is typically norminally 48V (The IEEE802.3af standard has set a maximum limit of 57V) and maximum available power to the device is around 13W (Hub cutoff current limit is typically around 350mA). The voltage used for PoE according IEEE 802.3af standard is within Safety Extra Low Voltage (SELV) limits. SELV is a protected secondary circuit designed to work under normal operating conditions and single fault conditions, such that its voltages do not exceed a safe value (60V max). The power can be fed to the cable from the power supplying special hub/switch or it can be fed with a separate power supply in the mid-span between the hub/switch and the power consuming device.

      Powered Device (PD) is a Terminal Device, which has been designed to receive power on the same cable as the data. Only one input connector is required. PD is the term defined by the IEEE 802.3af committee.

      Powered Sourcing Equipment (PSE) is an equipment supplying power over ordinary CAT5 cables. Power can be supplied on the spare or the data wires of the Ethernet LAN transmission cable. PSE is the term defined by the IEEE 802.3af committee.The devices are compatible with existing Ethernet or Fast Ethernet switches and enable a unified supply of data and power through a single connection by sending power over standard Category 5 twisted pair cable. The compatibility is done in such way that the power supplying device supplies the power only to devices that can handle this power (there is a standardized detection scheme for this).

      Some devices that are not originally designed for PoE Ethernet operation can be converted for PoE operation using a device called power splitter. Splitter is used to separate the power from the communication lines. This device is used when equipment has not been designed to receive the data and power input on the same cable. The Splitter splits the two components into two cables, one providing Ethernet data communication and the other providing power.

      As a result, enterprises can deploy IP telephony and other devices without the time and cost required to install separate power cabling to existing infrastructure. In addition, the Power Sorcing Equipment can be connected to a uninterruptable power supply (or 48V battery!) to ensure peripherals are powered inthe case of mains power failures. in this way Power over LAN provides continuous service during power outages by utilizing the same centralized UPS system that backs up the network. This architecture enhances a customer's investment in both Category 5 infrastructure and in Ethernet switch equipment. Managed Power over Ethernet eliminates the need to push a reset or power switch on remote, possibly difficult-to-reach PoE-powered devices. They can be turned on or off or reset by a network manager sitting at his or her desk. Power over Ethernet promises to enable these applications and many more by providing up to 12.95 watts of power (at 48 volts) over the same Category 5 cable that already delivers your standard 10/100/1000Mb Ethernet service.

      IEEE 802.3af using data pairs is 48 V DC protected power provided by the following wires in RJ-45 connector:

      • 1,2 = DC+
      • 3,6 = DC -

      IEEE 802.3af using spare pairs supplies 48 V DC protected power through the following RJ-45 connector pins:

      • 4,5 = DC +
      • 7,8 = DC -

      Many practical real-life equipment are designed in such way that they can accept the power from either one of those methods. And many equipment can also acccept power at any polarity.

      For safety reasons the isolation between the wiring and the equipment needs to be considered cafefully. Powered devices housed in non-conductive casing without any external connections except the Ethernet port do not require isolation. If the casing is conductive or other external connections exist, the isolated DC-DC power supply must be used (at least 1500 Volts AC insulation rating).

      Before the offical standard there has been properietray solutions that have used diferent pinouts and different voltages. Some systems have used some of their own equipment detection methods before supplying power, while some implementations (for example the simples DIY plans) have just supplies the power to the spare wires. With those systems that just supply power to wires, you should not that the power supplies to wires must be current limited / fuse protected (to avoid thin data wires from overhating if short circuit happens somewhere on the wiring) and there is risk of damaging some Ethernet equipment if plugged to output that supplies all the time power (that power can fry the termination resistors that quite many 100Mbit/s Ethernet equipment have connected to unused wire pairs to reduce emissions from the data wiring). Cisco's old pre-standard POE ethernet uses pinout 4,5=DC- and 7,8=DC+. Most other systems used 4,5 = DC + and 7,8 = DC - but voltages vary, usually being 12, 24 or 48V.

      The Internet Engineering Task Force (IETF) has been working in parallel with the IEEE to extend its Simple Network Management Protocol (SNMP) to apply to PoE ports as well. It has developed an Internet Draft that extends the Ethernet-like Interfaces MIB (RFC 2665) with a set of objects for managing Power Source Equipment and Powered Devices. IEEE 802.3af defines the hardware registers that would be used by a management interface. The IETF draft defines management data objects based on the information read from and written to these registers.

      Ethernet testing

      How Ethernet is depends on what level you want to test it. . The most basic test (a.k.a., "the fire test") is to connect a pair of devices to the network and see if they can communicate with each other. If you want to test the electrical integrity of the wiring (i.e., will it carry a signal properly), you need to use a suitable cable tester and/or TDR device for it. If you need to test the performance or troubleshoot protocol transmission problems, you will need special and usually very expensive software, usually coupled with custom hardware, to capture, optionally filter, and analyze the network packets.

      • Ethernet Test Suites    Rate this link
      • Measuring Ethernet Tap Capacitance - when a node is added to an Ethernet network which uses coaxial cabling, its nodal capacitance changes the impedance of the cable at the point of connection to the cable and impedance change causes a reflection of the Ethernet waveform, IEEE802.3 standard specifies a maximum value of capacitance that a node may add to the network, as well as a minimum node to node distance spacing    Rate this link

      Special techniques

      • Can I change the MAC address of a NIC? - Some NICs have the ability to change the MAC (Media Access Control) address through software. If you NIC and driver support this, Windows 2000 can change it using the tips shown in this article.    Rate this link
      • Designing for EMC in Datacom Systems - Electromagnetic compliance is not an area of "black magic," but rather another aspect of engineering where systematic design procedures, along with a little educated debugging, can provide quality solutions and shorten the design cycle. By following simple design rules, a two-layer, 10-Mbps Ethernet PCB can be created with minimal effort in debug time.    Rate this link
      • Magic Packet - Technology Application in Hardware and Software - Magic Packet technology is used to remotely wake up a sleeping or powered off PC on a network. This is accomplished by sending a specific packet of information, called a Magic Packet frame, to a node on the network. This application note addresses the use of the Magic Packet technology in conjunction with green PC hardware and system-level software. Its objective is to assist individuals in using this new technology in their own environments.    Rate this link

    Fast Ethernet (100 megabit/s)

    Fast Ethernet standard defines the 100-Mbps Fast Ethernet system which operates over twisted-pair and fiber optic media. The Fast Ethernet specifications include mechanisms for Auto-Negotiation of the media speed. This makes it possible for vendors to provide dual-speed Ethernet interfaces that can be installed and run at either 10-Mbps or 100-Mbps automatically.

    There are three media varieties that have been specified for transmitting 100-Mbps Ethernet signals.

    • The "T4" segment type is a twisted-pair segment that uses four pairs of telephone-grade twisted-pair wire. This is not much used.
    • The "TX" segment type is a twisted-pair segment that uses two pairs of wires and is based on the data grade (Category 5) twisted-pair physical medium. This is the most widely used version. The 100BASE-TX system operates over two pairs of wires (unshielded or shielded), one pair for receive data signals and the other pair for transmit data signals. The most popular wiring used today is Category 5 unshielded twisted-pair cable.
    • The "FX" segment type is a fiber optic link segment based on the fiber optic physical medium standard developed by ANSI and that uses two strands of fiber cable.
    The TX and FX medium standards are collectively known as 100BASE-X. The 100BASE-TX Ethernet segments are defined as link segments in the Ethernet specifications. A link segment is formally defined as a point-to-point medium that connects two and only two devices. A typical installation uses multiport repeater hubs, or packet switching hubs, to provide a connection between a larger number of link segments. You connect the Ethernet interface in your computer to one end of the link segment, and the other end of the link segment is connected to the hub. That way you can attach as many link segments with their associated computers as you have hub ports, and the computers all communicate via the hub. This means that the physical topology supported by twisted-pair link segments is the star. In this topology a set of link segments are connected to a hub, radiating out from the hub to the computers like the rays from a star.

    The 100BASE-TX specifications allow a segment of up to 100 meters. Two 100 meter 100BASE-TX segments can be connected together through a single Class I or Class II repeater. This provides a system with a total diameter of 200 meters between two communicating devices. If longer distances are needed, an Ethernet switch is needed in between (this breaks the network to two part with their own 200 meter limits).

    In 100BASE-T, the signal is encoded using a 125 MHz clock. 100BASE-TX standard adopts a 3-level from of data encoding called MLT-3. Here, the output (encoded) signal is selected from a repeating 4-state pattern of {1, 0, -1, 0}. If the next data bit is a 1, the output transitions to the next state in the pattern. If the next data bit is a 0, the output remains constant. This method of data encoding has the advantage that the highest frequency in the encoded signal occurs when transmitting a long sequence of data bit 1's, in which case the encoded signal repeats the {1, 0, -1, 0} pattern, which has a cycle length of 1/4 of the basic clock rate. Thus, in this worst case the primary energy component would be at 32.5 MHz when using a 125 MHz clock. For other data bit patterns, the energy would be distributed at lower frequencies. The data is 100Mbit/s Ethernet data is encoded using 4B-5B encoding before passing it to MLT-3 coder (this gives the 125 Mbit/s coding rate). The coding for fiber optic communications is somewhat simpler. The data is 100Mbit/s Ethernet data is just encoded using 4B-5B encoding before passing it to fiber (this gives the 125 Mbit/s coding rate). The 4B5B code first popularized by FDDI and then 100BASE-FX codes 4 bits of data to 5 bits of code. Of the 32 possible 5-bit code groups, 4B5B coding selects 16 combinations, plus a few control codes, that have either two or three bits set to one. With random data, the long-term average number of ones and zeros is the same; however, in the short term, you may receive an arbitrarily long string of code groups that have only two bits set to one. The dc content of such a signal is 2/5, not 1/2, which leads to a dc-offset error in an ac-coupled receiver of 10%. This DC-offset results in a receiver noise-margin penalty of 2 dB compared to ideal system. 4B5B transmit one additional symbol for every 4 bits of useful data, amounting to an overhead of 25%. 100Base-TX Ehternet system can be cosidered as an syncronous network, and there is always some signals flowing between the devices connected to different ends of cable to keep them in sync.

    100Base-TX uses the same pinout as 10Base-T:
    Pin #Signal NameFunction
    1TD+Transmit Data
    2TD-Transmit Data
    3RD+Receive Data
    4NCNo Connection
    5NCNo Connection
    6RD-Receive Data
    7NCNo Connection
    8NCNo Connection

    Many modern 100Base-TX implementations are built in sych way that TX and RX pairs may be interchanged (auto-detection circuitry manages to get the system working). The pins marked here with "no connection" are not used for actual data transmission. Usually those not used wires are terminated to reduce the RF emissions from the cable and for improving cable transfer characteristics. For designers of Ethernet systems, the Bob Smith Termination technique is regarded as an important approach for migrating potential RF emission and susceptibility problems for Ethernet infrastructure. A ?Bob Smith? termination is often provided for the unused signal pairs of the twisted-pair interface (RJ-45 pins 4, 5, 7, and 8) and the media-side center taps. This circuit is used to enhance EMI and ESD performance of the system. There are many variations on this technique (one common is 50 ohm resistor to each signal wire, the signals from one pair connected together, and those two interconnection points then connected toghether with two 50 ohm resistors). The Bob Smith termination can be broken down into two circuits. One circuit provides termination for the unused signal pairs of the twisted-pair interface. The unused pairs are connected together through a 75Ω impedance matching circuit, and then to chassis ground through a 0.001 ?F, 2kV capacitor. The capacitor provides a discharge path on the unused pairs. The second circuit provides termination for the media-side center taps and is comprised of individual 0.001 uF, 2kV capacitors to chassis ground. Separate capacitors are used for the receive and transmit center taps.

    The 100baseTX technology that uses two wire pair is the mainstream Fast Ethernet technology. In the early years of Fast Ethernet there were also two other competeting technologues: 100baseVG and 100baseT4. The 100baseVG and 100baseT4 protocols use four pairs, allowing the 100Mbit/s traffic to run even on lower quality than CAT5 wiring (CAT3/CAT4 cables also worked well). Your chance of running into those four pair Fast Ethernet variants is extremely small nowadays.

    Along with 100baseTC Fast Ethernet system cabe Ethernet auto-nagotiation system. When ethernet was first designed, negotiation didn't exist: there was nothing to negotiate because everything has CSMA/CD 10 Mbit/s (half-duplex). With introduction of Fast Ethernet the number of possibilities increased to different combinations of operation speed (10/100 Mbit) and operation mode (classical half-duplex or newer full-duplex). Many moden switches have also added option to automatically detect/nagotitate which one pair of wires is used for TX/RX (allows operation with both direct and cross over cables directly). How Autonagotiation is supposed to work:

    • When both sides of the link are set to autoneg, they will "negotiate" the duplex setting and select full duplex if both sides can do full-duplex.
    • If one side is hardcoded and not using autoneg, the autoneg process will "fail" and the side trying to autoneg is required by spec to use half-duplex mode.
    • If one side is using half-duplex, and the other is using full-duplex, sorrow and woe is the usual result.
    So, the following table shows what will happen given various settings on each side:
                     Auto       Half       Full
    
       Auto        Happiness   Lucky      Sorrow
    
       Half        Lucky       Happiness  Sorrow
    
       Full        Sorrow      Sorrow     Happiness
    
    Happiness means that there is a good shot of everything going well. Lucky means that things will likely go well, but not because you did anything correctly. Sorrow means that there will be a duplex mis-match (thigns do not work well).

    When there is a duplex mismatch, on the side running half-duplex you will see various errors and probably a number of late collisions. On the side running full-duplex you will see things like FCS errors. Note that those errors are not necessarily conclusive, they are simply indicators.

    Most 100 Mbit/s equipment use Auto-neg. Most times it works wery well and without problems. At some rare cases the auto-neg system does not work well with all equipment. In those rare case it is a good idea to configure one end or both ends of the connection manually to some known working settings. Auto-neg is a required part of the gigabit ethernet standard.

    There are also applications were 100 Mbit/s Ethernet signals are transported through fiber optics. There are devices with fiber interfaces and transcivers with fiber connection. In those are quite rately seen in modern setups. Nowadays it is typical that the 100Mbit/s networks are generally wired with copper wires and use 100baseTX signaling format. On those rare situations where fiber is needed (for example for long distance links) it is quite typical that a media converter that converts between 100baseTX and the suitable fiber media (single mode or multi mode fiber) is used. There are converters tht can go even tens of kilometer in distance. The media converters are typically just signal converters that adapt the signal from edia to another (think like two ethernet transceivers with different interfaces back-to-back with very little electronics in between). For the long distance to work, the Ethernet devices connected to those media converters must work in full-duplex mode (the distance limit for 100 Mbit/s classic half-duplex CSMA/CD based Ethernet operation is 200 meters, for longer distances it does not work). A typical approach to for network builders that use media converters is to set the equipment connected to the media converters to work always on full-duplex mode, since the operation when devices are set to "auto" mode is typically not reliable (strange thing can happen when auto-nagotiation fails). If the switches are mangeable and are set to a fixed configuration everething is fine. As there is no autonegotiation defined on a 100BaseFX any possible negotiation is local to the switch and its adjacent converter. To be usefull with autonegotiating-only switches, most todays mdiaconverters negotitiate full-duplex unconditionally.

      Technical information

      With 100BaseT technology came the ability to perform auto-negotiation between each end of a 100BaseT connection. When the connection is established (plugging both ends of the UTP cable into their respective ports), a series of fast link pulses (FLP) are exchanged between the ports. The 33 pulses contain 17 clock pulses and 16 data pulses. The 16 data pulses form a 16-bit code indicating the capabilities of the port, such as Communication mode (half duplex or full duplex) and speed (10, 100, 10/100). Fast Ethernet contains specifications for two types of repeaters, Class I and Class II. Class I repeaters are slower (140 bit times for its round-trip delay) than Class II repeaters (92 bits times or less), but provide functions such as translation between the many different 100BaseT technologies. Class II repeaters, although faster, support only a single technology.Standard topologies for 100BaseT networks are one Class I repeater, which provides a network diameter of 200 m using copper cable and stations that may be 100 m from the repeater, and two Class II repeaters. The latter are connected via a 5-m cable that provides a diameter of 205 m and stations that may be 100 m from each repeater. The 100BASE-TX and 100BASE-FX media standards used in Fast Ethernet are both adopted from physical media standards first developed by ANSI, the American National Standards Institute. The ANSI physical media standards were originally developed for the Fiber Distributed Data Interface (FDDI) LAN standard (ANSI standard X3T9.5), and are widely used in FDDI LANs.

      Full duplex Ethernet flow control

      Flow control is a mechanism created to manage the flow of data between two full-duplex Ethernet devices. Through flow control, a device that is oversubscribed - either macroscopically from a system resource perspective or microscopically on a port-by-port basis - sends a pause message to its link partner to temporarily reduce the amount of data it's transmitting. Otherwise, buffer overflow occurs, data is lost and retransmission is required.

    10 Gigabit Ethernet

    10GBE is a new 10 gigabit version of Ethernet. Standardization got ready at the end of year 2002. 10GBE is so fast that it works only on fiber optic links.

    10GBE is designed to give more bandwidth in metro, access, and transport systems. In order to service two broad network applications, the IEEE is defined a separate 10 Gigabit Ethernet WAN physical (PHY) layer and 10 Gigabit Ethernet LAN PHY.

    The LAN PHY is intended to maximize the data rate to 10 Gbps cheaply for short distances, while the WAN PHY is rate compatible with the existing OC-192 (9.95328 Gbps) WAN infrastructure. The data rate and frame structure for the WAN PHY were specifically engineered to match current SONET/SDH WAN and optical networking data rates. This was done so that 10 Gigabit Ethernet traffic could be format-and rate-compatible with existing SONET/SDH and optical transport infrastructure. Rate matching is required to accommodate a rate of 10 Gbps at the media access control (MAC) and the WAN PHY running at 9.953-Gbps line rate (MAC adds extra spaces between frames in LAN implementation to match WAN speed).

    The LAN PHY data rate is chosen to operate at 10 Gbps to optimize for throughput. In order to facilitate longer reach applications, additional fiber management capability has been added. The line rate of the LAN PHY depends on the coding scheme employed. The serial LAN PHY uses 64B/66B coding, while in applications using 4-l optics, 8B/10B is used. 8B10B transmit one additional symbol for every 4 bits of useful data, amounting to an overhead of 25%. 64B65B substantially reduce the required overhead, and thus the number of symbols you need to transmit, but at the expense of simplicity and low latency. The WAN PHY employs a basic SONET frame and scrambling to transport Ethernet data. The 64B/66B code characters generated from the same coding scheme used by the LAN PHY are encapsulated in SONET frames rather than being directly fed to the optics. Frame delineation within the received SONET payload is accomplished by recognizing valid 64B/66B data blocks.

    • 10 Gigabit Ethernet - Convergence of LAN and WAN - This is a a silide set from BICSI 2000 Fall Conference, Nashville, TN.    Rate this link
    • 10-Gigabit Ethernet takes on SONET - The 10-Gigabit Ethernet specification promises not only to eventually enable delivery of Gigabit Ethernet to the desktop but also to bridge the gap between datacomm and telecomm.    Rate this link
    • Manning Up for 10 Gigabit Ethernet - Emerging Ethernet technology promises more bandwidth in metro, access, and transport systems.    Rate this link
    • Son of Gigabit Ethernet - 10-Gigabit Ethernet is coming soon to a LAN (and MAN) near you    Rate this link
    • The Once and Future Ethernet - The inevitable move to 10-Gb speeds will make Ethernet the architecture of choice for large-scale networks.    Rate this link
    • 10GBASE-T PHYs Pose Challenges to Designers - Moving some of the signal processing tasks to the analog front end of the PHY could hold the key for the effective development of transceivers for emerging copper-based 10GE design.    Rate this link
    • The slow road to 10-Gbps Ethernet - Lower speeds don't usually equate with innovation. But in the realm of 10-Gbps Ethernet, two "low-speed" interconnect standards, LX4 and CX4, are reducing implementation cost and enabling new applications. Network engineers are not using 10GE links solely for long-haul applications. They're also installing them within facilities such as data centers. In fact, more than 50% of the 10GE ports being consumed are finding use in local, short-reach links. LX4 implements 10GE with each of the four lanes operating at 3.125 Gbps. Spec-compliant LX4 modules will achieve the 802.3ae limit of 240 to 300m with great margin. In fact, vendors such as Emcore have demonstrated links as long as 2 km on legacy MMF. CX4 targets the shortest of connections using copper wiring. CX4 permits as much as 15m of 24 AWG cable.    Rate this link
    • Breaking conventional wisdom: 10-Gigabit Ethernet fiber costs less than copper - New generations of optical modules are more compact, lower power and lower cost than previous generations. Higher levels of integration among the modules combined with silicon photonics technology, have the potential to bring optical interconnects into the same price range that has been the exclusive domain of copper, promoting the adoption of 10 Gigabit Ethernet.    Rate this link

    HomePNA

    HomePNA is a phoneline networking standard which allows usingnormal telephone line wiring for LAN wiring inside home. The Home Phoneline Networking Alliance (HomePNA) is an association of leading companies working together to help ensure adoption of a single, unified phoneline networking industry standard and rapidly bringing to market a range of interoperable home networking solutions.

    Currently actively used standard versions are HomePNA 1.0 and HomePNA 1.1.Those are also sometimes marketed with name HPNA. They provide 1 Mbit/s networking over ordinary home telephone wiring up to 150 meters. The card products for this look very much like Ethernet cards (just cost somewhat more). For getting good compatilibitly with differnet manufacturersthere is "Home Phoneline Network Certified" marking on manyproducts which are are known to nicely interwork with otherproducts marked with that mark. This HomePNA versions 1.0 and 1.1 can coexist on the same linewith technologies like analogue telephone (PSTN) and ADSL. It means that standard telephones, V.90 (56K) and other dial up analog modems, faxes and answering machines, as well as DSL service can be used simultaneously with HomePNA because, even though they exist on the same telephone wires, they occupy different frequency bands. HomePNA utilizes 4-10 MHz frequency range while analogue telephone and ADSL use lower frequencies (15Hz-3kHZ for PSTN and 25kHz-1.1MHz for ADSL).

    The HomePNA technology is a usual Ethernet with 1 Mbit/s (HomePNA 1.0) and 10 Mbit/s (HomePNA 2.0) in all aspects. The CSMA/CD, IEEE-802.3, MAC addresses are applicable not only for the Ethernet but also for both HomePNA standards. This technology differs from the Ethernet only on a physical level. The HPNA 1.0 technology5 uses a pulse position modulation (PPM) technique with a spectral efficiency of 0.16 bits/baud, resulting in a 1-Mbps data rate. HPNA selected the 4- to 10-MHz band for several reasons. The lower limit of 4 MHz makes it feasible to implement the filters needed to reduce out-of-band interference between HPNA and splitterless ADSL. After modeling several thousand representative networks with capacitive telephones and common wire lengths, it was determined that the spectrum above 10 MHz was much more likely to have wider and deeper nulls caused by reflections.7 Cross talk between phone lines increases with frequency. The particular choice of 4 to 10 MHz only overlaps a single amateur radio band (40 meters), which simplifies ingress and egress filtering.

    And installation of HomePNA cards doesn't differ from that of HomePNA adapters and similar Ethernet produicts. Operating systems operate with these adapters as with usual Ethernet ones. All HomePNA adapters must be connected to the same phone line. In there are muliple telephone lines (telephone numbers) in a home, all of the HomePNA adapters must be plugged into the same telephone line. The HomePNA devices use the middle two conductors of the telephone jack (RJ11).

    In some DSL installations, filters, micro-filters or micro-splitters are provide for telephone devices. In-line filters are "low-pass" filters, which means that they allow the low frequency signals (voice) to pass through, while blocking the high frequency signals (data) from traveling through the phone cord to your telephone, fax or answering machine. DSL and HomePNA use the high frequency bandwidth of your telephone line to transmit and receive data. Telephones should be plugged generally into the phone jack on the HomePNA devices if possible. In many applications phones on the other outlets work nicely when just plugged in, but in some cases you could need a filter to filter out the HomePNA signal that tries to get to normal telephones. Some filters if plugged between the HomePNA device phone jack and the telephone may cause the HomePNA device to stop working. In some cases, it is helpful to block external signals from a HomePNA network (a filter between incoming telephone line and your home wiring).

    HomePNA 1.0 technology is basically just a usual Ethernet working at 1 Mbit/s and using different line coding. HPNA 1.0 uses PPM line coding. The HomePNA 1.0 standard doesn't adapt to quality of a line. If a packet is lost, it is to be resent. The distance where the real speed of 1 Mbit/s is present is about 150 m. But the increased power of the signal, which can be enabled or disabled in the hub, allows reaching 500 m. But in this case you will get more noise and pickups, that is why the decision of whether to enable it or not depends on a definite topology of the network. You can, however, adjust power of a signal separately on each port. Sometimes it should be done because of a growing number of collision (with large signals and at small distances), which affect a data rate. While the HomePNA 1.0 places a limitation on the quantity of working devices in such network (up to 25 devices), usage of hubs changes these limitations. HomePNA 1.1 is an extension to HomePNA 1.0 that provides the same 1 Mbit/s transmission technology and also a more robust transmission technology (700 kbit/s speed). Practically all modern HomePNA gear nowadays use HomePNA 1.1 specification.

    HomePNA 2.0 promises higher data speeds (10 Mbit/s or more)and transmission distances up to 320 meters. HomePNA Version 2.0 is designed to reach up to 1000 feet (300 meters) between any two adapters. If the network has more than two HomePNA adapters, all of the adapters must be within 1000 feet of each other. The actual distance may be greater or perhaps less depending on the type or wire, noise conditions and topology of the telephone wiring within your home. HomePNA 2.0 specification claims a line rate of 32 Mbps; however, when accounting for things like overhead and retransmission, the rate is closer to 20 Mbps. Due higher data rate, HomePNA 2.0 is more sensitive to telephone line crosstalk than earlier versions. HomePNA has completed version 3.0 of its spec in early June 2003. The technology offers a top data rate of 128 Mbits/sec and has deterministic QoS features. HomePNA 3.0 is reported to achieve data rates of more than 100 Mbits/sec in about 50 percent of the scenarios and more than 40 Mbits/sec in "just about all" the others. The first ICs for this specification were available at the end of year 2003. The specifications are 10 Mbit/s speed, a range of 350 m, the number of devices up to 32. Although the HomePNA 1 and 2 standards are compatible, the HomePNA 2.0 is based on different principles. It can adapt a data rate. HomePNA 2.0 network has been claimed operated flawlessly in such media as UTP 3 and 5, telephone cables, laminated metals, coaxial cables etc. The HomePNA 2.0 can also work good when we applied native signals to these cables - broadcasting, television, telephone etc. And within 350 m stipulated in the documents the speed didn't depends on a type of cable. The data rate is severely affected by different pickups. Try to coil a cable, and the signal may be lost. Everything depends on the equipment, especially when a signal goes via main telephone lines. HomePNA can run also though coaxial cables. In case of a coaxial cable the distance range is about 2.5-3 km. While HPNA 1.0 uses PPM, HPNA 2.0 uses quadrature amplitude modulation (QAM), both to get more throughput in the same bandwidth and to achieve greater robustness. However, because the channels may have very deep nulls, and multiple nulls in band, two techniques are used. The firsttechnique is adapting the modulation rate. Instead of having a fixed number of bits per symbol, a transmitter may, on a packet-by-packet basis, vary the packet encoding from 2 to 8 bits per symbol. A packet header is always encoded at 2 bits per symbol, so that every receiver can demodulate at least the packet?s header. The system uses a fixed 7-MHz carrier frequency and can operate at either 2 Mbaud or 4 Mbaud with modulation encodings of 2 to 8 bits per symbol. The base symbol rate is 2 Mbaud. At this rate, the system has a peak data rate ranging from 4 to 16 Mbps, though overhead reduces the actual throughput the system can achieve. In practice, to achieve performance equivalent to 10Base-T Ethernet, a packet must be sent at 6 bits per symbol. At its 2-Mbaud rate, HPNA 2.0 implements a modified version of QAM invented by Eric Ojard called frequency-diverse QAM (FDQAM). Because in FDQAM the baud rate is less than half the filter?s width, the output signal has two redundant copies of the baseband signal. Thus, the signal is frequency diverse, motivating the name FDQAM. FDQAM works robustly in many cases where uncoded QAM would fail. Such channels are common on home phone lines. In cases where the channel nulls are not particularly deep, HPNA 2.0 allows for a higher performance 4-Mbaud mode, which achieves peak data rates up to 32 Mbps and throughput above 20 Mbps. The frame begins with a known 64-symbol preamble. The preamble supports robust carrier sensing and collision detection, equalizer training, timing recovery, and gain adjustment. Following the preamble is a frame control field, the first part of which is an 8-bit frame type (=0 for IEEE 802.3 packet, other codes for future extenstions), followed by an 8-bit field that specifies the modulation format, followed by other miscellaneous control fields in frame control including an 8-bit CRC header. The remainder of the packet is exactly an 802.3 Ethernet frame followed by CRC16, padding, and EOF sequence. HPNA 2.0 is a CSMA/CD system, just like the standard IEEE 802.3 Ethernet. HPNA 2.0 introduces eight levels of priority and uses a new collision resolution algorithm called distributed fair priority queuing (DFPQ).

    The HomePNA has won a firm foothold on the market. First, the HomePNA 1.0 is used successfully in office buildings - practically all of them have their own telephone network, which can be used for the Internet as well. HomePNA networks can be built in those buildings which have phone jacks. I.e. you don't need hubs and switches, but only HomePNA cards.

    There are newer versions coming. Theoretically, the HomePNA 3.0 standard has every chance to reach 100 Mbit/s speed! HomePNA 3.0. Surpassing industry expectations, the final HomePNA 3.0 reaches an unprecedented data rate of 128 Mbps with optional extensions reaching up to 240 Mbps. As the only home networking industry specification capable of reaching above 100 Mbps and with inherent deterministic Quality of Service (QoS). HomePNA 3.0 greatly enhances version 2.0 capabilities adding deterministic QoS support for real-time data. The technology permits users to assign specific time slots for each stream of data guaranteeing that the real-time data will be delivered when it is required with predetermined latency and without interruption.

    There has been some standardization work on HomePNA technology in ITU also approved 2001 a set of standards for Home Phone-line Networking transceivers, ITU-T Recommendation G.989.1. This will allow home-networking devices (e.g. computer peripherals) to operate over existing telephone wiring. Newer ITU-T specifications include G.989.1, G989.2 and G989.3 based on the HomePNA 2.0 specification.

    Token Ring

    Token Ring is a network architechture which usestoken passing technology and ring type network structure.Token Ring is standardized in IEEE 802.5 standard.Token Ring was widely used competitior of Ethernet, butnowadays it's use has quite much faded to only thoseorganizations which have already large Token Ring infrastructure.


<[email protected]>

Back to ePanorama main page ??