Network Glossary

10Base2 – A variant of Ethernet, connecting stations via thin coaxial cable; maximum cable distance in one non-repeated segment is 185 meters.

10Base5 – A variant of Ethernet, connecting stations via thick coaxial cable; maximum cable distance in one non-repeated segment is 500 meters.

10BaseFL – A variant of Ethernet, connecting stations via fiber optic cabling.

10BaseT – A variant of Ethernet, connecting stations via twisted pair cabling.

100BaseFX – A variant of Ethernet which runs on multimode or single mode fiber optic cabling at 100 Mbps. This is one version of Fast Ethernet.

100BaseTX – A variant of Ethernet which runs on Category 5 unshielded twisted pair wiring at 100 Mbps. This is one version of Fast Ethernet.

1000Base-CX – A variant of Gigabit Ethernet which runs on twinaxial cable.

1000Base-LX – A variant of Gigabit Ethernet which runs on multimode and single mode fiber optic cable at a 1330 nm frequency.

1000Base-SX – A variant of Gigabit Ethernet which runs on multimode fiber optic cable at an 850 nm frequency.

1000Base-T – A variant of Gigabit Ethernet which runs on unshielded twisted pair cable.

802.x – The set of IEEE standards defining LAN protocols.

A

AAL – See "ATM Adaptation Layer".

ABR – See "Available Bit Rate".

Access Control Method – This is the main distinguishing feature between different LAN technologies. It regulates each workstation's physical access to the cable (transmission medium), and determines the order in which nodes gain access so that each user gets efficient service. Access methods include token passing, which is used in token ring and FDDI, and Carrier Sense Multiple Access with Collision Detection (CSMA/CD), which is employed by Ethernet and Fast Ethernet.

Active Monitor – A node on a token ring network which purges the ring and generates a new token (when necessary), initiates and monitors neighbor notification, and maintains the master clock.

Address Mask – Used to select bits from an Internet address for subnet addressing. The mask is 32 bits long and selects the network portion of the Internet address and one or more bits of the local portion. Sometimes called Subnet Mask.

Address Resolution Protocol (ARP) – Because host addresses and protocols vary in length and value, they are often incompatible with the corresponding 48-bit Ethernet address. The Address Resolution Protocol (ARP) allows for the dynamic distribution of the information needed to build tables which facilitate the translation of an incompatible address into a 48-bit Ethernet address. This protocol has been defined by the IETF.

Adjusted Ring Length (ARL) – Calculated to ensure that if there is a ring failure, the longest ring path is still within specifications. Generally associated with token ring, Adjusted Ring Length ensures that the secondary ring can still function properly in the event of a failure on the shortest trunk cable.

Agent – The portion of the system in the client-server model that performs information preparation and exchange on behalf of a client or server application.

American National Standards Institute (ANSI) – A U.S. standards body. ANSI is a member of the International Organization for Standardization (ISO).

Applications Program Interface (API) – Software designed to make a computer's facilities accessible to an application program. All operating systems and network operating systems have APIs. In a networking environment it is essential that various machines' APIs are compatible, otherwise programs would be exclusive to the machines in which they reside. As networking has developed, some APIs have become de facto standards, including NetBIOS and DOS 3.1.

ARQ – See "Automatic Repeat Request".

ASIC (Application-Specific Integrated Circuit) – A chip designed for a specific application, generally by the manufacturer of the product in which the chip is used.

Asynchronous – A method of transmitting data whereby each byte is clocked separately. One start bit is added to the beginning, and one or more stop bits to the end, of each character. Asynchronous transmission is the most rudimentary form of data communication, as the originating and recipient machines do not have to be in sync. It is commonly used for low-speed transmission, as with a PC's serial port. This meaning of the term "asynchronous" is completely different from that in the next definition.

Asynchronous Transfer Mode (ATM) – A high-speed, connection-oriented switching and multiplexing technology for transmitting information across a wide area or local area network. ATM divides information into fixed-length cells capable of transmitting different types of traffic simultaneously, including voice, video, and data.

ATM – See "Asynchronous Transfer Mode".

ATM Adaptation Layer (AAL) – Provides a conversion function to and from ATM for various types of information, including voice, video, and data. There are several versions of AAL, each applicable to a given information type. All of them convert elements of an information stream (such as voice frame and data packets) into cells, giving ATM the versatility to carry many different types of data, from constant-rate voice data to highly bursty messages generated by LANs, all within the same cell format.

ATM-ARP – Resolves MAC to ATM address translation.

ATM Forum – An international consortium of hundreds of companies and users chartered to accelerate the use of ATM products and services by developing specifications and promoting the technology. The ATM Forum is not a de jure standards body, but on a de facto basis it has been responsible for development of a wide range of ATM standards. It works in cooperation with standards bodies such as ANSI and ITU, submitting to them proposed standards.

ATM LAN Emulation (LANE) – See "LAN Emulation".

Attachment Unit Interface (AUI) – Defined in the IEEE 802.1 specification as the interface between an Ethernet MAU and DTE. Basically, the way an Ethernet station connects to a transceiver sitting on a thick Ethernet cable.

Attenuation – The progressive weakening of a signal as it travels away from its point of origin.

AUI – See "Attachment Unit Interface".

Authentication – A means to establish or prove identity; verifying eligibility of users, machines, or objects.

Authorization – Privileges granted and resources available.

Automatic Repeat Request (ARQ) – A type of error correction ensuring that a transmitting device automatically resends any data containing errors.

Autonomous System – Internet (TCP/IP) terminology for gateways (routers) that fall under one administrative entity and cooperate using a common Interior Gateway Protocol (IGP). See "Subnet".

Available Bit Rate (ABR) – A form of ATM transmission in which an information stream is allowed to access the bit rate left after the predictive and guaranteed service traffic (CBR / VBR) are served. ABR provides a dynamically negotiated rate and includes a congestion control capability. A typical use is for support of workstations equipped directly with ATM network interface cards.

B

Backbone – LAN or WAN connectivity between subnets across a high-speed network. Often applied to a high-speed campus network, such as ATM OC-12 or Gigabit Ethernet, that interconnects lower-speed networks, such as ATM OC-3 or Fast Ethernet. Fiber optic cable is often used.

Backplane – Describes the bus or matrix that traditionally resides at the back of a modular networking product, and into which the modules are plugged.

Bandwidth – (1) The range of signal frequencies that can be carried on a communications channel. The capacity of a channel is measured in cycles per second, or hertz (Hz), between the highest and lowest frequencies. (2) Commonly, the carrying capacity of a digital translation facility, measured in bits per second (bps).

Baseband – A technique whereby digital input is directly applied to transmission media without the intervention of a modulating device. Baseband is generally applied in an environment with high bandwidth over a short distance. It is generally considered easier and more cost-effective than broadband. Ethernet, token ring, FDDI, and ATM generally use baseband.

Basic Rate Interface (BRI) – An ISDN subscriber interface which operates over a single copper cable connection, providing one control (D) channel at 16 Kbps, and one or two bearer (B) channels, at 64 Kbps each. The two B channels are sometimes combined to provide a single 128 Kbps service. BRI is the interface commonly provided to residential ISDN subscribers.

Bellcore (Bell Communications Research) – Telecommunications research and development organization currently owned by the seven U.S. regional Bell operating companies.

BootP (Bootstrap Protocol) – A UDP/IP-based protocol that allows a booting host to configure itself dynamically, and more significantly, without user supervision. It provides a means to assign a host its IP address, a file from which to download a boot program from a server, that server's address, and (if present) the address of an Internet gateway.

Border Gateway Protocol (BGP4) – Interdomain policy routing protocol for communications between a router in one autonomous system (AS) and routers in other AS's.

Bridge – See "MAC-Layer Bridge".

Broadband – Characteristic of any network that multiplexes multiple, independent carrier signals onto a single cable. This is usually accomplished through frequency division multiplexing. Broadband technology allows several signals to coexist on a single cable; traffic from one signal does not interfere with traffic from another, since data is transmitted on a different frequency. Cable television uses broadband.

Broadband ISDN (B-ISDN) – The new generation of Integrated Services Digital Network (ISDN) which carries digital data, voice, and video over SONET networks. B-ISDN allows Asynchronous Transfer Mode (ATM) and Synchronous Transfer Mode (STM) services to operate on the same network.

Broadband LAN – A LAN which uses frequency division multiplexing (FDM) to divide a single physical channel into a number of smaller, independent frequency channels. The different channels created by FDM can be used to transfer different forms of information, such as, voice, data, and video.

Broadcast – A packet delivered to all workstations on a network. Broadcasts exist at layer two and at layer three.

Broadcast Domain – The set of end stations which receive the same broadcast packets.

Broadcast Storm – An overload condition in a network created by an incorrect packet broadcast onto the network that causes multiple hosts to respond all at once. Typically the response contains equally incorrect packets, which causes the storm to grow exponentially in severity.

Broadcast and Unknown Server (BUS) – An ATM LANE process which relays broadcast and multicast packets, and packets with unknown destination addresses, to all Emulated LAN clients. It can be implemented on any ATM device, including a file server, a switch, an access device, or a router.

Bus – (1) A conductor, or set of conductors (e.g. wires), that serves as the interconnection between a related set of devices. (2) A specific type of backplane in which all slots are connected to a common set of wires or traces on which they send to and receive from all other slots. (3) A network topology in which the signals sent by one device are received by all other devices. Each device then selects those transmissions addressed to it based on address information contained in the transmission.

BUS – See "Broadcast and Unknown Server".

C

CAC – See "Connection Admission Control."

Cache – A special area of high-speed memory used to store addresses in a switch. Also called a forwarding table.

Campus Network – A network which covers a single customer location, such as a building, a floor of a building, or all of the buildings on a large commercial, educational, or other campus.

Carrier Sense Multiple Access with Collision Detection (CSMA/CD) – A contention-based network access method in which any computer may attempt to communicate at any time. Since there is no centralized force controlling the medium, a device must first sense whether or not the medium is in use. If the medium is unused the device then transmits. If two computers sense that a channel is open and transmit at the same time, the result is a collision, after which there is a random pause determined individually by each transmitting machine. Each machine then senses the line again and, if it is available, retransmits.

CBR – See "Constant Bit Rate".

CE – See "Circuit Emulation".

Cell – A fixed-length transmission unit which forms the basis of ATM. Each cell is 53 bytes in length, divided into a 48-byte payload and a 5-byte header.

Cell Discard – The process within an ATM switch of discarding cells when the switch’s buffer capacity is exceeded.

Cell Loss Priority (CLP) – A one-bit field in the ATM cell header that determines whether or not a given cell should be dropped by network equipment during periods of congestion. This explicit loss priority can be set by the source node or by the network. A CLP which equals zero receives high network priority while a CLP which equals one is dropped during periods of congestion.

Cell Loss Ratio (CLR) – The ratio of discarded cells to cells that are successfully transmitted. Specifically, CLR equals the number of discarded cells divided by the number of transmitted cells.

Checksum – A computed value which is the outcome of a mathematical function applied to the contents of a packet. This value is sent along with the packet when it is transmitted. The receiving system computes a new checksum based on the received data and compares this value with the one sent with the packet. If the two values are the same, the data was received correctly.

Circuit Emulation (CE) – A service provided across a public or private ATM network which emulates the characteristics of a leased-line service.

Circuit Switching – A communications method whereby a circuit is held open and maintained only while the sender and recipient are communicating. This is different from a dedicated circuit which is held open regardless of whether data is being sent or not, and different from a datagram / connectionless network, in which data flows without the establishment of a connection.

Classless Inter-domain Routing (CIDR) - An IAB protocol which uses variable-length subnetting techniques to distribute the allocation of Internet address space. CIDR is needed to address the exhaustion of class B network address space, the growth of Internet routing tables, and the eventual exhaustion of the 32-bit IP address space.

CLP – See "Cell Loss Priority".

CLR – See "Cell Loss Ratio".

Coaxial Cable (Coax) – Formerly common in Ethernet networks, coax comes in various types with different transmission characteristics. It is copper-based, with an inner conductor surrounded by an outer conductor, with insulation between the two, insulation around the outer conductor, and a jacket. Coax is less flexible than twisted pair cable, but more resistant to EMI and physical breakage.

Collapsed Backbone – A network architecture in which a router or switch provides a building or campus backbone using a star topology.

Collision – Concurrent Ethernet transmissions from two or more devices on the same segment. A collision is sensed by the transmitting stations as an over-voltage condition, and they retransmit after waiting a random amount of time.

Common Open Policy Service (COPS) - An IAB client/server model for supporting policy control over QoS signaling protocols with similar properties as ReSerVation Protocol (RSVP). In RSVP, the router, or network device, must respond to bandwidth reservation requests; with COPS, the router forwards the bandwidth request to the nearest COPS policy server. The server makes the end-to-end bandwidth decision; the router implements it. The result is less overhead on the router and overall lower network latency.

Congestion Control – Mechanisms that control traffic flow so that intermediate network devices and end stations are not overwhelmed. Used in connection-oriented networks such as frame relay and ATM. More sophisticated mechanisms are needed to deal with congestion in large networks carrying different types of traffic. Sometimes referred to as flow control.

Connection Admission Control –The set of actions taken by the network during the call setup phase (or during call re-negotiation phase) in order to determine whether a connection request can be accepted or should be rejected (or whether a request for re-allocation can be accommodated).

Connection-Oriented – The model of interconnection in which communication proceeds through three well-defined phases: connection establishment, data transfer, connection release. Examples of connection-oriented networks include ATM, frame relay, X.25, and Internet TCP.

Connectionless – The model of interconnection in which communication takes place without formal connection establishment. Examples include Ethernet, Internet IP, and UDP.

Constant Bit Rate (CBR) – A form of ATM transmission in which a fixed bit rate is provided, with clock frequency and phase maintained end-to-end. Typical uses include emulation of a leased-line circuit, and carrying traditional 64 Kbps PCM voice.

CRC – See "Cyclical Redundancy Check".

CSMA/CD – See "Carrier Sense Multiple Access with Collision Detection".

Customer Service Unit (CSU) – A device used at the customer premise to connect a device, such as a PBX, to a public digital network facility, such as a T1 line. Provides repeater and control functions.

Cut-through – (1) A form of switching, typically LAN switching, in which the switch begins to forward the initial portion of a packet to its destination while the remainder of the packet is still being received. This was useful when the throughput of LAN protocols was highly degraded by latency in the data path. It is uncommon today. (2) A form of switching, typically in an ATM network, in which a routing process is used to set up a connection between two devices, but the data subsequently flows directly between the two devices, without passing through the routing process. MPOA is one important example.

Cyclical Redundancy Check (CRC) – An error-checking mechanism for layer-two data transmissions. Polynomial calculations are performed using only the number of bits in the message. The bits are then sent along with the data to its recipient. The recipient checks the data it receives and repeats the calculation. If there are any discrepancies between the results of the two calculations, the recipient requests the originator to resend the data.

D

DAS – See "Dual Attached Station".

Data Communications Equipment (DCE) – Traditional data communications terminology for the equipment that enables a DTE to communicate over a telephone line or data circuit. The DCE establishes, maintains, and terminates a connection as well as performing the conversions necessary for communications.

Data Link Connection Identifier (DLCI) – A unique number assigned to a PVC end point in a frame relay network. Identifies a particular PVC end point within a user's access channel in a frame relay network, and has local significance only to that port.

Data Terminal Equipment (DTE) – Traditional data communications terminology for a device receiving and/or originating data on a network. Typically a computer or dumb terminal.

Datagram – A self-contained, independent entity of data carrying sufficient information to be routed from its source to the destination computer without reliance on earlier exchanges between the source, the destination computer, and the transporting network.

DCE – See "Data Communications Equipment".

Decryption – The inverse of "encryption."

DHCP – See "Dynamic Host Configuration Protocol".

Digital Service Unit (DSU) – A device used at the customer premise to connect a data device, such as a computer, to a public digital network facility, such as a T1 line. Provides electrical translation and line coding. Technically, this is generally a DSU / CSU, combining both functions.

Digital Signature – Electronic means to ensure message integrity, typically based on a public key cryptosystem.

Distance Vector Multicast Routing Protocol (DVMRP) – A protocol designed to support the forwarding of multicast datagrams through an internetwork. DVMRP constructs source-rooted multicast delivery trees using variants of the Reverse Path Broadcasting (RVP) algorithm. Some version of DVMRP is currently deployed in the majority of MBONE routers.

DNS – See "Domain Name System".

Domain – In networking, a subdivision of the hosts on a network. The division can be physical, as in separate building LANs, or logical, as in giving the hosts in a particular administrative area their own group name even though they are on the same network.

Domain Name System (DNS) – An IAB standard that provides a globally-accessible table of domain names (e.g., xylan.com) and their corresponding IP addresses.

DS-0 – A 64 Kbps digital channel carried within a DS-1 or E1 signal.

DS-1 – The digital signal carried on a North American high-speed facility operating at 1.544 Mbps.

DS-3 – The digital signal carried on a North American high-speed facility operating at approximately 45 Mbps.

DS-3 – The 45 Mbps transmission rate carried on a US T3 facility.

DSU – See "Digital Service Unit".

DTE – See "Data Terminal Equipment".

Dual Attached Station (DAS) – A form of FDDI connection in which a dual counter-rotating ring is supported. Typically used for connecting concentrators and servers to a main ring.

Duplex – A technique allowing bi-directional, simultaneous transmission along a channel. Generally referred to as full duplex.

DVMRP – See "Distance Vector Multicast Routing Protocol".

Dynamic Host Configuration Protocol (DHCP) – A protocol within the TCP/IP family which allows a server process to assign a layer-three (IP) address to a device, when the device requests it. DHCP replaces static configuration of IP addresses by network operators, and in some cases can substantially simplify network management.

Dynamic Routing – A procedure for sending messages across a network by which line failure or overload results in message rerouting.

E

E-1 – The digital signal carried on a high-speed facility operating outside of North America at 2.048 Mbps.

E-3 – The digital signal carried on a high-speed facility operating outside of North America at approximately 34 Mbps.

Early Packet Discard (EPD) – An intelligent cell discard process that occurs within an ATM switch when its buffer capacity is exceeded. EPD discards all cells that originated as members of a single data frame, since the entire frame would have to be retransmitted if even one cell were discarded.

EFCI – See "Explicit Forward Congestion Indication".

EGP – See "Exterior Gateway Protocol".

ELAN (Emulated LAN) – See "LAN Emulation".

Electro-Magnetic Interference (EMI) – Electrical interference with the operation of an electrical device, or with a communications transmission, caused by magnetic radiation. Typically originates from another electrical device, or from a communications transmission in a nearby cable.

Encapsulation – The technique used by layered protocols in which a layer adds header information to the protocol data unit (PDU) from the layer above.

Encryption – The process of converting information from an easily understandable format (plain text) into apparent random gibberish (ciphertext) by the use of well-defined rules and calculations known as algorithms or cipher. A process used to ensure the privacy and confidentiality of information. The reverse process is decryption.

Ethernet – The most common layer-two protocol used in LANs. Ethernet is a 10 Mbps CSMA/CD standard originally developed by Xerox to run on thick coaxial cabling. It has evolved and now runs primarily on twisted pair cabling.

Explicit Forward Congestion Indication (EFCI) – EFCI is an indication in the ATM cell header. A network element in an impending-congested state or a congested state may set EFCI so that this indication may be examined by the destination end-system. For example, the end-system may use this indication to implement a protocol that adaptively lowers the cell rate of the connection during congestion or impending congestion. A network element that is not in a congestion state or an impending congestion state will not modify the value of this indication. Impending congestion is the state when network equipment is operating around its engineered capacity level.

Explicit Rate – Explicit Rate is a type of flow control mechanism defined by the ATM Forum Traffic Management 4.0 standard. With Explicit Rate flow control ATM-attached sources are periodically issued resource management (RM) cells which stipulate a maximum cell rate (measured in cells per second) at which the device can transmit and be guaranteed that traffic will not be discarded by the network.

Exterior Gateway Protocol (EGP) – A routing protocol used by gateways in two-level Internets. EGP is used in the Internet core system.

F

Fabric Blocking – The state that can exist within a switch when its internal switching fabric is not capable of handling simultaneous maximum-rate transmissions by all inputs.

Fast Ethernet – A version of Ethernet which operates at 100 Mbps. See 100BaseTx and 100BaseFX.

Fault-Tolerance – The ability of a device to prevent or recover from network and internal failures. Key elements of fault tolerance include hot-swappable modules, redundant load-sharing power supplies, passive backplanes, and redundant cooling systems.

FDDI – A local area network based on a backbone of dual counter-rotating 100 Mbps fiber optic rings. One of the rings is normally designated as the primary ring; the other is the secondary ring. The dual ring is connected to single-attached "slave" rings through concentrators.

FDM – See "Frequency Division Multiplexing".

Fiber Channel – A form of high-speed fiber optic transmission designed primarily for communications between mainframe computers, and between mainframe computers and high-speed peripherals such as disk drives. Sometimes used for general-purpose networking.

Field Programmable Gate Array (FPGA) – A general-purpose semiconductor component which can be customized to operate physically as though it were a chip dedicated to a specific task.

File Transfer Protocol (FTP) – The protocol within the TCP/IP protocol suite which is used to transfer files between computers.

Firewall – A security mechanism which protects a server, a subnet, or an entire end user location from unauthorized access. Firewalls can be standalone devices, or they can be incorporated into routers and switches.

Flooding – Transmission of a frame to all devices in a segment or ring (in routed networks) or a virtual LAN (in a virtual LAN-based network). Flooding is performed on broadcasts, multicasts, and frames whose destination address is unknown.

Flow Control – See "Congestion Control".

Forwarding Table – A special area of high-speed memory used to store addresses in a switch. Also called a cache.

FPGA – See "Field Programmable Gate Array".

Fragmentation – The process in which a protocol data unit is broken into smaller pieces to fit the requirements of a network. The reverse process is reassembly.

Frame – A unit of information in a layer-two protocol. In LANs, a frame is a MAC-layer unit containing both control information and an entire layer-three packet. Although the term "packet" is sometimes used to mean a frame, the term "frame" is never used to describe a layer-three packet.

Frame Relay – An ITU standard for the interface to a public frame-switching network designed to provide high-speed frame transmission with minimum delay across the wide area. It operates at layer two, and is used in public and private networks, gradually replacing X.25 and leased-line networks.

Frame Tagging – A process of adding a header to the front of a layer-two frame, so that additional information needed to manage the frame through the network is provided. This information can include membership in one or more virtual LANs, priority information, and / or quality of service information.

Frequency Division Multiplexing (FDM) – Method by which the available transmission frequency range is divided into narrower bands; each of these bands is used for a separate channel. This allows several signals to be sent over the same transmission medium.

FRF.5 – A Frame Relay Forum specification for internetworking ATM and frame relay networks. FRF.5 allows ATM networks to transparently pass frame relay data link connection indentifiers over ATM virtual channel indentifiers. This allows ATM networks to act as high-speed backbones for frame relay networks.

FRF.8 – A Frame Relay Forum specification for service internetworking ATM and frame relay. This allows frame relay data link connection indentifiers to be directly mapped into ATM virtual channel indentifiers. FRF.8 allows frame relay devices to directly communicate with ATM attached devices.

FRF.9 – A Frame Relay Forum specification for data compression within frame relay.

FTP – See "File Transfer Protocol".

Full-Duplex – A communications method in which a transmission path is provided in each direction, so that each end can simultaneously transmit and receive.

G

Gateway – A combination of hardware and software that interconnects otherwise incompatible networks or networking devices. The term is sometimes used to indicate a device (uncommon now) which translates between disparate protocol stacks.

Gbps – Billions of bits per second.

Gigabit Ethernet – A variant of Ethernet which operates over multimode fiber optic cable, single mode fiber optic cable, or unshielded twisted pair, at 1,000 Mbps.

H

Half-Duplex – A communications method in which one end transmits while the other receives, then the process is reversed. This was common in wide area point-to-multipoint circuits, such as those used in many SNA networks.

Head End – A central point in a broadband network that receives signals on one set of frequency bands and retransmits them on another set of frequencies. The head end is viewed as a central hub. Every transmission on a broadband network must go through the head end.

Header – A portion added to the beginning of a message containing essential information such as the source address, destination address, and control information.

Head-of-Line Blocking – The state that exists when frames or cells within a single input queue are destined for multiple outputs, and one output is congested, thus delaying all cells.

HEC (Header Error Control) – An 8-bit Cyclic Redundancy Code (CRC) computed on all fields in an ATM header; capable of detecting single-bit and certain multiple-bit errors. HEC is used by the physical layer for cell delineation.

Horizontal Cabling – That portion of a building’s cabling system which extends from the wiring closets to the individual workstations, servers, telephones, and other devices. This is generally copper twisted pair cable.

Hot Standby Router Protocol (HSRP) – A Cisco-driven IAB protocol that allows hosts to appear to use a single router and to maintain connectivity even if the actual first hop router fails. Multiple routers participate in this protocol by creating the illusion of a single virtual router. The protocol ensures that one and only one of the routers is forwarding packets on behalf of the virtual router. End hosts forward their packets to the virtual router. See also "Virtual Router Redundancy Protocol."

HTML – A form of page description language used in the World Wide Web.

HTTP (Hypertext Transfer Protocol) – An IAB protocol used in the World Wide Web which defines how requests for HTML and graphics files which make up a WWW page are handled between the web server and the client browser.

Hub – The center of a star topology network or cabling system. Typically used in older Ethernet and token ring networks. A device connected to a hub receives all the transmissions of all other devices connected to that hub. Hubs are now being replaced in many cases by LAN switches.

Hybrid Network – A LAN consisting of a number of topologies and access methods. For example, a network that includes both token ring and Ethernet.

I

IAB – See "Internet Activities Board".

ICMP – See "Internet Control Message Protocol".

ICMP Router Discovery – An extension of the IAB Internet Control Message Protocol (ICMP) that enables hosts attached to multicast or broadcast networks to discover the IP addresses of their neighboring routers without requiring static default route configurations.

IDF – See "Intermediate Distribution Frame".

IEEE – See "Institute of Electrical and Electronic Engineers".

IEEE 802.1D – See "Spanning Tree".

IEEE 802.1p – An IEEE standard for prioritizing time-critical flows and filtering multicast traffic to contain traffic in layer-two networks. The 802.1p header includes three bits for prioritization, allowing for eight priorities to be established.

IEEE 802.1Q – An IEEE standard for providing a virtual LAN capability within a campus network, used in conjunction with IEEE LAN protocols such as Ethernet and token ring.

IEEE 802.2 – A data link standard outlining how basic data connectivity over cable should be set up. Used with the IEEE 802.3, 802.4 and 802.5 standards.

IEEE 802.3 – The IEEE's specification for Ethernet, including both physical cabling and layer-two protocol.

IEEE 802.5 – The IEEE's specification for token ring, including both physical cabling and layer-two protocol.

IEEE 802.10 – The IEEE’s protocol for providing security in a metropolitan area network. A variant of 802.10 has sometimes been used to provide a virtual LAN service within a campus network, although this is now generally replaced with 802.1Q.

IETF – See "Internet Engineering Task Force".

IGMP (Internet Group Management Protocol) – A protocol that runs between hosts and their immediately neighboring multicast routers; the mechanisms of the protocol allow a host to inform its local router that it wishes to receive transmissions addressed to a specific multicast group.

IGP – See "Interior Gateway Protocol".

IISP (Interim Interswitch Signaling Protocol) – An ATM Forum specification for signaling between ATM switches, using statically defined connections. Largely replaced by PNNI.

ILMI (Interim Local Management Interface) – An interim requirements definition in ATM Forum UNI 3.1. It supports bidirectional exchange of management information between UNI management entities related to the ATM layer and physical layer parameters.

Information Superhighway – A sadly meaningless phrase, generally associated with politicians, which implies something or other having to do with the Internet.

Institute of Electrical and Electronic Engineers (IEEE) – A standards-making body responsible for implementing many standards used in LANs, including the 802.x series.

Integrated Services Digital Network (ISDN) – A CCITT standard developed to cover a range of voice, data, and image services. It is intended to provide end-to-end, simultaneous handling of voice and data on a single link. Access channels include Basic Rate Interface (BRI) and Primary Rate Interface (PRI).

Intelligent Hub – A hub that adds network management capabilities, such as maintaining port statistics, determining port status, and automatically segmenting faulty ports. Also known as a second-generation hub.

Interior Gateway Protocol (IGP) – A type of protocol used to exchange routing information between collaborating routers on the Internet. RIP and OSPF are examples of IGPs.

Intermediate Distribution Frame (IDF) – In a structured building wiring system, the gathering point for cabling from a section of a building, such as a floor or a portion of a floor. Typically, multiple IDFs located in wiring closets connect to a central MDF.

International Organization for Standardization (ISO) – An international organization that develops standards. ISO is best known in networking for its seven-layer Open Systems Interconnection (OSI) reference model that conceptually organizes communications protocols into seven layers.

Internet Activities Board (IAB) – The technical body that oversees the development of the Internet suite of protocols.

Internet Control Message Protocol (ICMP) – The protocol used to handle errors and control messages at the IP layer. ICMP is actually part of the IP protocol.

Internet Engineering Task Force (IETF) – A body within the IAB which supports the development of new protocols for the Internet.

Internet Packet Exchange Protocol (IPX) – The layer-three protocol used in Novell’s NetWare protocol suite. IPX provides a connectionless datagram delivery service for transport-layer protocols such as SPX and NCP. Has nothing to do with the "Internet" as that term is commonly used today.

Internet Protocol (IP) – The layer-three protocol used in the TCP/IP set of protocols which support the Internet and many private networks. IP provides a connectionless datagram delivery service for transport-layer protocols such as TCP and UDP.

Internet ReSerVation Protocol (RSVP) – An IAB standard which allows an end device and a network to negotiate specific QoS characteristics.

Internetwork – Two or more networks connected by bridges or routers.

Intranet – The use of various Internet tools and protocols, especially HTTP and HTML, within an organization.

Inverse Multiplexing – The use of multiple circuits between two devices in which the circuits are treated as a single virtual channel. Traffic is spread across the circuits, and the loss of one circuit results in reduced bandwidth rather than loss of the connection.

IP – See "Internet Protocol".

IP Datagram – The fundamental unit of information passed across the Internet at layer three. It contains source and destination addresses along with the data, and a number of fields that define such things as the length of the datagram and the header checksum.

IP Switching – A form of layer-three cut-through switching pioneered by Ipsilon Corporation, which is now a division of Nokia. In IP Switching, the first packet, or packets, of each information flow are routed as in a traditional router-based network. However, if the routers detect that the flow is likely to be long-lived (as, for example, an FTP connection), then a cut-through path is set up between the end stations.

IPX – See "Internet Packet Exchange Protocol".

ISDN – See "Integrated Services Digital Network".

ISO – See "International Organization for Standardization".

ISO Reference Model for Open Systems Interconnection (OSI) – Developed by the International Standards Organization, the seven-layer model describing the process of network communication. It is intended to facilitate communications among computers from different manufacturers and to provide a common basis for coordinating international standards. Most modern protocols map to the OSI model to some extent, especially at the lower layers.

Isochronous – Signals which are dependent on some uniform timing or carry their own timing information embedded as part of the signal.

ITU (International Telecommunications Union) – An international body of member countries whose task is to define recommendations and standards relating to the international telecommunications industry. The fundamental standards for ATM have been defined and published by the ITU (previously CCITT).

J

Jitter – A short-term timing deviation.

K

Kbps – Thousands of bits per second.

L

Label Swapping – Also known as label switching. A general term for a layer-three switching mechanism which attaches a label, or tag, to each packet. The label provides intermediate switches with the information needed to forward the packet toward its destination.

LAN – See "Local Area Network".

LAN Emulation (LANE) – A set of protocols developed by the ATM Forum which allows legacy LAN protocols, such as Ethernet and token ring, and higher-layer protocols and applications which depend on LAN protocols, to work transparently across an ATM network. LANE translates address formats, emulates the LAN broadcast function, and automatically sets up ATM connections. LAN Emulation retains all Ethernet and token ring drivers and adapters; no modifications need to be made to Ethernet or token ring end stations. Multiple emulated LANs (ELANs) within the same ATM network are common. Also, single stations can belong to multiple ELANs.

LAN Emulation Client – An end device in a LANE application. Can be a workstation or server with an ATM NIC; more commonly, a LAN switch with an ATM uplink.

LAN Emulation Configuration Server (LECS) – A process within ATM LAN Emulation which assigns individual LAN Emulation Clients to emulated LANs.

LAN Emulation Server (LES) – A process within ATM LAN Emulation which translates between MAC addresses and ATM addresses.

Latency – Delay in a transmission path or in a device within a transmission path. Also referred to as propagation delay.

LDAP – See "Lightweight Directory Access Protocol".

Leased line – A transmission facility which is leased by an end user from a public carrier, and which is dedicated to that user’s traffic. Typically, frequency synchronization is maintained from one end of the circuit to the other. Leased line circuits are generally used less in recent times, while public data networks are more common.

LEC – See "LAN Emulation Client".

LECS – See "LAN Emulation Configuration Server".

LES – See "LAN Emulation Server".

Lightweight Directory Access Protocol (LDAP) – An IAB standard, based on the ITU X.500 standard, which provides a mechanism for communicating with a central directory that is shared by many different services. LDAP is likely to play a central role in managing dynamic networks.

LLC – See "Logical Link Control".

Lobe Port – In token ring, a port on a MAU or hub to which the cable from a device attaches. Lobe ports must receive a specific voltage from the attached device in order to allow the device into the ring.

Local Area Network (LAN) – (1) The network which interconnects all computing devices located within a single end user location; e.g., an integrated token ring / ATM network covering an entire campus. (2) A single layer-two network, which may be connected to other such networks within an end user location; e.g., a single Ethernet segment. To avoid confusing the two definitions, Xylan commonly refers to the former as a "campus" network.

Logical Link Control (LLC) – A sublayer of layer two which provides a connection between the layer-three protocol, such as IP or IPX, and the MAC layer protocol. LLC2, one form of LLC, provides a connection-oriented service.

Loopback – A testing method in which the transmitted data is looped back to the receiver.

M

MAC – see "Media Access Control".

MAC Address – The layer-two address of a LAN device.

MAC (Media Access Control) Layer – A sublayer of layer two that deals with the issues specific to a particular type of LAN; e.g., Ethernet or token ring.

MAC-Layer Bridge – A device used to forward data between LANs at layer two, by automatically filtering out traffic which is local to each LAN, while forwarding on traffic which is not local to each LAN. All broadcasts and multicasts, as well as all traffic with a destination address which has not been learned by the bridge, is forwarded.

MAC-Layer Protocol – See "Media Access Control".

MAC-Layer Switching – LAN data transferred through a network based on the source and destination addresses contained in the MAC header of the frame. Essentially the same as bridging, but almost always employing dedicated hardware to perform the switching.

Main Distribution Frame (MDF) – In a structured building wiring system, the central point for cabling throughout the building. Typically, multiple IDFs located in wiring closets connect to a central MDF.

MAN – see "Metropolitan Area Network".

Management Information Base (MIB) – A database of objects that can be accessed via a network management protocol. See "SNMP."

MAU – See "Media Access Unit".

Maximum Lobe Length (MLL) – The maximum allowable distance between a node and a MAU or hub on a token ring network.

Maximum Transfer Unit (MTU) – An IAB discovery protocol that polls the network for the highest MTU possible between a source and a destination. The result is an optimized frame size that prevents fragmentation and yields better end-to-end throughput.

Mbps – Millions of bits per second.

MDF – See "Main Distribution Frame".

Media Access Control (MAC) – The way in which LAN workstations share access to a transmission medium. MAC-layer protocols include Ethernet, token ring, and FDDI. Has absolutely nothing to do with the Apple Macintosh computer.

Media Access Unit – In token ring, a hub which interconnects the devices connected to the ring, and in turn connects to other MAUs through Ring In / Ring Out connections. Generally a MAU is not managed via software.

Metropolitan Area Network – A network spanning a geographical area greater than a LAN, but less than a WAN.

MIB – See "Management Information Base".

Microsegmentation – The process of dividing up LAN segments to contain fewer users on a shared media LAN, increasing performance by reducing congestion. It is generally implemented with LAN switches.

Mid-Level Networks – The transit networks that make up the second level of the Internet hierarchy. They connect the sub-networks to the backbone networks. Also known as regionals.

MLL – See "Maximum Lobe Length".

MPOA – See "Multi-Protocol Over ATM".

MTU – See "Maximum Transfer Unit".

Multicast – A form of broadcast in which a packet is delivered to a pre-defined subset of all possible destinations. A specific multicast destination address is used.

Multilink PPP – A form of PPP which uses inverse multiplexing of multiple wide area circuits to achieve a higher-bandwidth virtual connection.

Multimode – A form of fiber optic cabling in which light is able to follow multiple paths as it traverses the cable. Less expensive, and with a lower maximum rate and distance, than single mode fiber optic cable.

Multiplex – To transmit two or more messages or message streams on a single channel, typically through the use of frequency-division multiplexing, time division multiplexing, or statistical time division multiplexing.

Multiplexer – A device used for division of a transmission facility into two or more subchannels, either by splitting the frequency band into narrower bands or by allotting a common channel to several different transmitting devices one at a time. Also known as a mux.

Multi-Protocol Over ATM (MPOA) – A protocol developed by the ATM Forum which provides a standard method for the routing of multiple protocols across an ATM network. The first version of MPOA supports only IP traffic.

N

NAT – See "Network Address Translation".

NDIS (Network Driver Interface Specification) – Developed by Microsoft for writing hardware-independent drivers. NDIS allows multiple protocol stacks (e.g., TCP/IP and NetWare) to share a single network interface module and the software which supports it.

NEBS (Network Equipment-Building System) – Bellcore has devised a three-tier system of criteria for NEBS compliance to ensure that the telecommunications equipment that various operating companies purchase is suitable for their needs, and to reduce the time and expense for manufacturers. The main purpose of this three-tier system is to identify criteria levels and the impact of any non-conforming result. The levels cover safety, environmental, and equipment operability under increasingly rigorous conditions.

NetWare – A protocol suite developed by Novell Corporation. The second most widely used protocol in LANs, after TCP/IP.

Network Address Translation (NAT) – A process by which addresses (typically IP addresses) are translated from one set of addresses to another. Typically used to allow a large address space to be used within a campus network when only a very limited address space is available for that organization’s connection to the Internet.

Network Segment – A portion of a network set apart from other network sections by a bridge, router, or switch. Each network segment supports a single medium access protocol.

NHRP (Next Hop Resolution Protocol) – An IAB protocol which provides a cut-through service between end stations in an ATM network.

NIC (Network Interface Card) – A physical plug-in module which goes into a workstation or server and provides the connection to a network.

Non-Real Time Variable Bit Rate (nrt-VBR) – A form of ATM transmission in which clock frequency can vary, but mean variation of delay between cells is guaranteed. A typical use is transmission of stored video.

O

OC-3 – A standard ATM / SONET rate and framing specification; approximately 155 Mbps.

OC-12 – A standard ATM / SONET rate and framing specification; approximately 622 Mbps.

ODI (Open Datalink Interface) – The Novell standard for hardware-independent drivers. ODI can simultaneously support multiple protocol stacks.

Open Shortest Path First (OSPF) – An IAB protocol which is used by IP routers to determine the optimal path along which to move a packet. Like other routing protocols, OSPF requires regular exchange of information among the routers, from which each router calculates the optimal path toward any given subnet. OSPF is a relatively advanced "vector" protocol.

Open Systems Interconnection (OSI) – See "International Organization for Standardization".

Optical Bypass – A capability in FDDI for enhanced failure resistance. A DAS station, such as a concentrator, generates a DC voltage to an attached mechanical optical bypass unit, through which pass all optical signals between the station and the ring. If the station fails, the voltage drops, and the optical bypass unit defaults to a state in which the ring optically passes straight through the bypass unit and cuts out the station.

OSPF – See "Open Shortest Path First".

P

Packet – (1) A variable-length layer-three protocol entity containing address and control information, plus data. Examples include IP and IPX packets. (2) A variable-length layer-two protocol entity containing address and other control information, plus data. Examples include Ethernet and token ring packets. These are also referred to as "frames," and in this book the term "packet" generally refers to a layer-three entity.

Packet Filtering – The ability of a bridge, router, or gateway to limit propagation of packets between two or more interconnected networks.

Packet Switching – A communications method in which variable-length packets are individually routed between hosts.

Partial Packet Discard (PPD) – A process of intelligent cell discard that occurs in an ATM switch when its buffer capacity is exceeded. PPD discards traffic for whole upper-layer PDUs when congestion is encountered. This is done by identifying which cells have been segmented from an individual frame (or packet) and discarding those cells associated with that frame.

PCR – See "Peak Cell Rate".

PDU – See "Protocol Data Unit".

Peak Cell Rate (PCR) – The maximum rate at which ATM cells can be transmitted across a virtual circuit, specified in cells per second and defined by the interval between the transmission of the last bit of one cell and the first bit of the next.

Permanent Virtual Circuit (PVC) – A connection in a connection-oriented network which is established through configuration, rather than dynamically.

Phase Jitter – The result of repeaters regenerating a signal which has experienced envelope delay in transmission through electronics and cable. Phase jitter is removed by processing the data stream through a buffer and reclocking it.

PNNI (Private Network-to-Node Interface) – An advanced, dynamic routing protocol that operates between ATM switches. It is based on link-state protocols, such as OSPF, with extensions that enable switches to advertise their own capabilities, such as capacity and delay.

Point-to-Point Protocol (PPP) – The successor to SLIP, PPP is a layer-two protocol which provides router-to-router and computer-to-network connections across a wide area circuit, generally in a TCP/IP network. See "SLIP."

Port Mirroring – A capability, typically in a switch, which allows a network manager to replicate the real-time data flow from one port at another port. Typically, the second port is attached to a protocol analyzer.

PPD – See "Partial Packet Discard".

PPP – See "Point to Point Protocol".

pps – Packets per second.

Primary Rate Interface (PRI) – An ISDN subscriber interface which operates over a copper or fiber cable connection, providing one control (D) channel at 64 Kbps, and 23 (North America) or 30 (international) bearer (B) channels, at 64 Kbps each. The B channels are sometimes combined to provide various transmission rates. PRI is the interface commonly provided to business ISDN subscribers.

Protocol – A formal description of messages to be exchanged and rules to be followed for two or more systems to exchange information.

Protocol Converter – A device for translating the protocol of one network or device to the corresponding protocol of another network or device. A protocol converter enables equipment with different conventions to communicate with one another.

Protocol Data Unit (PDU) – A defined data unit passed from one protocol layer to another. Each protocol layer encapsulates the PDU from the layer above within the information which it adds.

Protocol Independent Multicast (PIM) Dense Mode – An IAB multicast protocol similar to DVMRP in that it uses Reverse Path Forwarding but does not require any particular unicast protocol. It is useful when multicast senders/receivers are in close proximity to one another, there are few senders and many receivers, the volume of multicast traffic is high, and the stream of multicast traffic is constant.

Protocol Independent Multicast (PIM) Sparse Mode – An IAB multicast protocol that works by defining a rendezvous point that is common to both sender and receiver. Sender and receiver initiate communication at the rendezvous point, and when flow begins it occurs over an optimized path. This is useful when there are few receivers in a group, senders and receivers are separated by WAN links, and the traffic is intermittent.

Proxy – The mechanism whereby one system acts for another system in responding to protocol requests.

Proxy ARP – The technique in which one device, usually a router, answers and issues ARP requests for another device.

PVC – See "Permanent Virtual Circuit".

Q

Quality of Service (QoS) – The requirements a network must provide to an individual flow of information (e.g., a voice call, an interactive video conference, a data file transfer) so that the information is optimally delivered. Elements of QoS can include maximum transit delay, maximum variability of delay, level of data path security, and prioritization with regard to other traffic.

R

RADIUS (Remote Access Dial-In User Service) – An IAB UDP-based protocol used for carrying authentication, authorization, accounting, and security information between a client and a server. Developed to better manage large serial line and modem pools, RADIUS leverages a single user database containing user ID/password and user authorized server types. The client/server model supports security via PAP, CHAP, UNIX login, and other authentication schemes, such as challenge/response systems.

Random Early Discard (RED) – A process of intelligent cell discard that occurs within an ATM switch when its buffer capacity is exceeded. RED discards cells in a round-robin fashion among affected connections.

Real Time Variable Bit Rate (rt-VBR) – A form of ATM transmission in which clock frequency can vary, but maximum delay and maximum variation of delay between cells are guaranteed. A typical use is real-time videoconferencing.

Remote Monitor (RMON) – An IAB specification for a set of MIBs which are used to communicate statistical, event, and other management control information between a managed device and a network management station.

Repeater – A device which propagates electrical signals from one segment to another without routing, buffering, or filtering.

Request for Comment (RFC) – A document written and registered within a process of dialogue managed by the IAB; the collective substance of the most recent RFPs on various topics relating to the Internet forms the body of Internet standards.

RIF – See "Routing Information Field".

Ring – A LAN topology in which each device is connected to two other workstations, with the connections forming a ring. Data is sent from device to device around the ring in a single direction. Each device acts as a repeater by resending messages to other devices. Examples include token ring and FDDI.

Ring Error Monitor for Token Ring – A ring resident function which maintains statistical records of error conditions on the ring operation.

Ring In and Ring Out (RI / RO) – The token ring connectors on the MAU that connect it to other MAUs. Unlike lobe ports, Ring In / Ring Out ports support a "wrap" capability; if an RI / RO cable is disconnected, the ring wraps back on itself, maintaining viability.

RIP – See "Routing Information Protocol".

Riser Cabling – That portion of a building’s cabling system which extends from the main distribution frame to the wiring closets. For data, this is often fiber optic cable. For voice, it is fiber optic cable if the PBX is distributed, and twisted pair copper cable otherwise.

RJ-11 – A standard connector commonly used to terminate voice connections.

RJ-45 – A standard connector commonly used to terminate data connections.

RMON – See "Remote Monitor".

Round Trip Delay – A measure of the delay in a network from request sent to reply received.

Route – The path that network traffic takes from its source to its destination.

Router – A layer-three device responsible for making decisions regarding which of several paths network traffic will follow. To do this, it uses a routing protocol to gain information about the network, and algorithms to choose the best route based on several criteria (known as routing metrics). Routers interconnect subnets.

Routing – The process of delivering a message across a network or networks via the most appropriate path.

Routing Domain – A set of routers exchanging routing information within administrative boundaries.

Routing Information Field (RIF) – A field in a token ring or FDDI frame header which provides information used by source-routing bridges to move the frame through a network. The RIF specifies a series of interleaved ring numbers and bridge numbers.

Routing Information Protocol (RIP) – An IAB protocol which is used by IP routers to determine the optimal path along which to move a packet. Like other routing protocols, RIP requires regular exchange of information among the routers, from which each router calculates the optimal path toward any given subnet. RIP is a relatively simple "link-state" protocol.

rt-VBR – See "Real Time Variable Bit Rate".

RSVP – See "Internet ReSerVation Protocol".

S

SAP – See "Service Advertising Protocol".

SAR – See "Segmentation and Reassembly".

SAS – See "Single Attached Station".

Segment – An electrically continuous piece of a bus-based LAN, typically Ethernet. Segments can be joined together using repeaters, switches, bridges, or routers.

Segmentation – Increasing the available bandwidth per device by dividing a network with bridges, switches, or routers to decrease the number of nodes on a segment.

Segmentation and Reassembly (SAR) – A process that occurs within an ATM access device, such as a LAN switch, or sometimes in a LAN switch. In a SAR process, information carried in data frames, such as Ethernet, or voice frames, such as a DS-0 channel, is divided into cells. The SAR is responsible for mapping data from the AAL Convergence Sublayer into the cell payloads of an ATM cell stream.

Sequenced Packet Exchange (SPX) – The layer-four protocol used in Novell’s NetWare protocol suite. SPX provides a connection-oriented transport-layer service.

Service Advertising Protocol (SAP) – A protocol used in Novell’s NetWare protocol suite which allows servers to inform workstations of their availability, through periodic broadcast packets.

Shielded Twisted Pair (STP) – Copper cable that includes one or more sets of cable pairs which have been molded into an insulating material and covered by a braided shielding conductor. STP offers better noise protection than unshielded twisted pair (UTP) but is much more expensive and more difficult to use. Commonly associated with early token ring networks.

Signaling – Communications between devices to set up calls and tear them down.

Simple Network Management Protocol (SNMP) – An IAB protocol designed to manage networking devices. With SNMP a management station can configure a supported device (SET); can request that the device send statistical, status, and configuration information (GET); and can receive unsolicited alarms from a device (TRAP).

Single Attached Station (SAS) – A form of FDDI connection in which a single ring is supported. Typically used for connecting workstations and servers to a concentrator.

Single Mode – A form of fiber optic cabling in which light follows a single path as it traverses the cable. More expensive, and with a higher maximum rate and distance, than multimode fiber optic cable.

SNA (Systems Network Architecture) – An important protocol suite developed by IBM Corporation beginning in the 1970s for use in both local and wide area communications. Pioneered many modern communications techniques. Many SNA networks are still in place at large organizations, although they are generally converting to TCP/IP.

SNMP – See "Simple Network Management Protocol".

SONET (Synchronous Optical Network) – A set of standards for data communication over fiber optic cable at speeds of 51.84 Mbps and above.

Source Route Bridge – A bridge which is capable of processing the Routing Information Field in a token ring or FDDI frame to determine whether or not to forward that particular frame.

Source Routing (SRB) – A protocol in which the end stations determine the path that frames will follow between them. An end station sends a preliminary route-finding broadcast frame, which turns into many frames, each following a separate route, and each accumulating a statement of the path it has followed. The one that arrives first is assumed to have followed the fastest path, and its path is then specified in all subsequent frames. Source routing is used in some, but not all, token ring and FDDI networks.

Source Route Transparent (SRT) – A protocol which is used in some token ring networks, which uses source routing for frames that need it, and uses transparent bridging for other frames. A variant (SRTB) translates from one type of frame to the other, so that end stations with disparate configurations can communicate.

Spanning Tree – A protocol specified in the IEEE 802.1D standard which allows a network to have a topology that contains physical loops. Spanning Tree operates in bridges and switches. It opens certain paths to create a tree topology, thereby preventing packets from looping endlessly on the network.

Spanning Tree Domain – A portion of a network in which a single Spanning Tree operates.

SPX – See "Sequenced Packet Exchange".

S/T Interface – A physical interface in an ISDN Basic Rate service which uses two copper pairs.

Standby Monitor – Any 802.5 token ring adapter currently attached (active) to the ring which is not the active monitor. One standby monitor assumes the role of the active monitor if it is no longer present on the ring.

Star – A network topology in which each node is connected to a central point.

Station Cabling – See "Horizontal Cabling".

Statistical Time Division Multiplexing (STDM) – Also known as statistical multiplexing. A form of time division multiplexing in which a given data stream can obtain more or less bandwidth dynamically, based on its needs and on the demands of other data streams. Widely used in devices such as routers, LAN switches, and frame relay switches.

Store and Forward – A method of switching in which a message is received as a whole, buffered, and then resent. All routers and virtually all current switches work in this manner. See "Cut-through".

STP – See "Shielded Twisted Pair". Not related to the popular engine-cleansing fuel additive.

Subnet – A portion of a network in which all stations share a common subnet address.

Subnet Address – The subnet portion of an IP address.

Subnet Mask – See "Address Mask".

Sustainable Cell Rate – The maximum throughput bursty traffic can achieve within a given virtual circuit without risking cell loss.

SVC – See "Switched Virtual Circuit".

Switched Virtual Circuit (SVC) – A connection in a connection-oriented network which is established dynamically, rather than through network configuration. An SVC is set up through a protocol which operates between a switch and an end station, and between switches.

Synchronous – Signals that are sourced from the same timing reference and have the same frequency. For example, in high-speed wide area digital communications, the network commonly provides a reference clocking source to which each subscriber’s equipment synchronizes its transmissions.

Synchronous Transfer Mode – B-ISDN communications method that transmits a group of different data streams synchronized to a single reference clock.

T

T1 – See "DS-1".

T3 – See "DS-3".

Tagging – See "Frame Tagging".

TAXI – An early standard for ATM transmission at 100 Mbps. Not commonly used now.

TCP – see "Transmission Control Protocol".

TCP/IP – The various protocols which support the Internet and many private networks. An instance of the advantages of synergistic cooperation over central planning.

TELNET – The protocol within the TCP/IP protocol suite which provides a terminal emulation function.

Time Division Multiplexing (TDM) – A method of multiplexing in which multiple information streams "take turns" with a single communications channel. Each stream is allocated a specified percentage of the common channel.

TLS – See "Transparent LAN Service".

Token – A unique packet that is passed around a token ring or FDDI LAN continuously. When a device wishes to transmit, it waits until it receives the token, attaches its message to the token, and transmits it. The device then removes its message from the ring when the token and message return to it.

Token ring – A network architecture standardized in IEEE 802.5 in which the devices on a ring transmit data while they are in possession of a token which passes from node to node continuously. Token ring operates at 4 or 16 Mbps.

Topology – Can be either physical or logical. Physical topology describes the physical connections of a network and the geometric arrangement of links and nodes that make up that network. Logical topology describes the possible logical connections between nodes, and indicates which pairs of nodes are able to communicate.

TOS – See "Type of Service".

TP/PMD – See "Twisted Pair / Physical Medium Dependent".

Transmission Control Protocol (TCP) – A layer-four protocol in the set of protocols which supports the Internet and many private networks. TCP provides a guaranteed transport service.

Transparent LAN Service (TLS) – A service provided by a common carrier in which multiple end user locations are interconnected, using layer-two or layer-three processes, in such a way that the entire network appears to be located on a single site. Commonly implemented today using an ATM service, with LAN switches equipped with ATM uplinks at the customers’ sites.

Tree – A LAN topology in which there is only one route between any two of the nodes on the network. The pattern of connections resembles a tree.

Twisted Pair – Insulated copper wires twisted together with the twists or lays varied in length to reduce potential signal interference between the pairs. They are usually bundled together and wrapped in a cable sheath. New data grade Unshielded Twisted Pair (Category 5) is specified for 100 Mbps transmission.

Twisted Pair / Physical Medium Dependent (TP/PMD) – A physical-level specification for FDDI which allows it to operate over unshielded twisted pair and shielded twisted pair copper cable. Sometimes referred to as "CDDI".

Type of Service (TOS) – A field within an IP header which can be used by the device originating the packet, or by an intermediate networking device, to signal a request for a specific QoS level. 1

78177176175174

U

U Interface – A physical interface in an ISDN Basic Rate service which uses a single copper pair.

UBR – See "Unspecified Bit Rate".

UDP – See "User Datagram Protocol".

UNI (User-to-Network Interface) – An interface point between ATM end users and a private ATM switch, or between a private ATM switch and the public carrier ATM network; defined by physical and protocol specifications per ATM Forum UNI documents. The standard adopted by the ATM Forum to define connections between users or end stations and a local switch.

Unicast – A frame which is sent from one station to another. A unicast contains the specific MAC addresses of the source and destination devices.

Unspecified Bit Rate (UBR) – A form of ATM transmission in which an information stream is supported on whatever bandwidth is available after other connection types have been satisfied. No congestion control is provided. UBR is commonly used to support information streams originating in LAN switches with ATM uplinks.

URL (Universal Resource Locator) – The English equivalent of an IP address and path that describes the location of an HTML (or other type) document on the World Wide Web. The first part of the URL describes the protocol to be used, the second is the DNS location of the server where the document is located, and the last is the path to the document.

User Datagram Protocol – A layer-four protocol in the TCP/IP protocol suite which serves as a connectionless alternative to TCP. Among other functions, UDP is used by SNMP.

V

VCI – See "Virtual Channel Identifier".

Virtual Channel – A single connection across a UNI or NNI allowing the switching of various ATM cells in a virtual path to different destinations.

Virtual Channel Identifier (VCI) – Identifier in an ATM cell of local significance across UNI or NNI which distinguishes data of one virtual channel from the data of another.

Virtual Circuit – A link that behaves like a dedicated point-to-point line or a system that delivers packets in sequence, as would happen on an actual point-to-point network.

Virtual Path – Contains virtual circuits that are to be switched together to a common destination such as an inter-exchange carrier.

Virtual Path Identifier (VPI) – The field in the ATM cell header that labels (identifies) a particular virtual path.

Virtual Router Redundancy Protocol (VRRP) – A non-Cisco-driven IAB protocol that allows several routers on a multi-access link to utilize the same virtual IP address. One router will be elected as a master, with the other router(s) acting as backup(s) in case of master router failure. Host systems may be configured with a single default gateway, rather than running an active routing protocol. See also Hot Standby Router Protocol.

VLAN (Virtual Local Area Network) – In a switched network, a logical collection of devices, such as all the workstations and servers with a particular IP subnet address, which are grouped into a broadcast domain.

VPI – See "Virtual Path Identifier".

W

Web – See "World Wide Web".

Wide Area Network (WAN) – A network which covers a larger geographical area than a single end user location, and in which telecommunications links are implemented, normally leased from service provider(s).

World Wide Web (Web) – The set of HTTP servers, and clients which access them, which are interconnected via the Internet.

X

X.25 – An ITU standard for the interface to a public packet-switching network. Generally connection oriented.