Home | Important Glossary in Computer Networks

Chapter: Computer Networks

Important Glossary in Computer Networks

Computer Networks - Important Glossary in Computer Networks




3DES: Triple DES, a version of DES that uses three keys, effectively increasing the key size and robustness of the encryption.


3G: Third-generation mobile wireless, a class of cellular wireless technologies based on CDMA.


4B/5B: A type of bit-encoding scheme used in FDDI, in which every 4 bits of data are transmitted as a 5-bit sequence.


802.3: IEEE Ethernet standard. 802.5: IEEE token ring standard.


802.11: IEEE wireless network standard. 802.17 IEEE resilient packet ring standard.


822: Refers to RFC 822, which defines the format of Internet email messages. See SMTP. AAL: ATM Adaptation Layer. A protocol layer, configured over ATM. Two AALs are defined for data communications, AAL3/4 and AAL5. Each protocol layer provides a mechanism to segment large packets into cells at the sender and to reassemble the cells back together at the receiver.


ABR: (1) Available bit rate. A rate-based congestion-control scheme being developed for use on ATM networks. ABR is intended to allow a source to increase or decrease its allotted rate, based on feedback from switches within the network. Contrast with CBR, UBR, and VBR. (2) Area border router. Router at the edge of an area in a link-state protocol.


ACK: An abbreviation for acknowledgment. An acknowledgment is sent by a receiver of data to indicate to the sender that the data transmission was successful.


additive increase/multiplicative decrease: Congestion window strategy used by TCP.


TCP opens the congestion window at a linear rate, but halves it when losses are experi enced due to congestion. It has been shown that additive increase/multiplicative decrease

is a necessary condition for a congestion-control mechanism to be stable.

AES: Advanced Encryption Standard. A cryptographic cipher that has been proposed to supersede DES.


AF: Assured forwarding. One of the per-hop behaviors proposed for Differentiated Services.


ALF: Application Level Framing. A protocol design principle that says that application programs better understand their communication needs than do general-purpose transport protocols.


AMPS: Advanced mobile phone system. Analog-based cell phone system. Currently being replaced by digital system, known as PCS.


ANSI: American National Standards Institute. Private U.S. standardization body that commonly participates in the ISO standardization process. Responsible for SONET.

API: Application programming interface. Interface that application programs use to access the network subsystem (usually the transport protocol). Usually OS-specific. The


socket API from Berkeley Unix is a widely used example.

area: In the context of link-state routing, a collection of adjacent routers that share full routing information with each other. A routing domain is divided into areas to improve scalability.

ARP: Address Resolution Protocol. Protocol of the Internet architecture, used to translate high-level protocol addresses into physical hardware addresses. Commonly used on the Internet to map IP addresses into Ethernet addresses.


ARPA: Advanced Research Projects Agency. One of the research and development organizations


within the Department of Defense. Responsible for funding the ARPANET as


well as the research that led to the development of the TCP/IP Internet. Also known as DARPA, the D standing for Defense.


ARPANET: An experimental wide-area packet-switched network funded by ARPA and begun in the late 1960s, which became the backbone of the developing Internet.


ARQ: Automatic repeat request. General strategy for reliably sending packets over an unreliable link. If the sender does not receive an ACK for a packet after a certain time period, it assumes that the packet did not arrive (or was delivered with bit errors) and retransmits it. Stop-and-wait and sliding window are two example ARQ protocols. Contrast with FEC.


ASN.1: Abstract Syntax Notation One. In conjunction with BER, a presentationformatting standard devised by the ISO as part of the OSI architecture.


ATM: Asynchronous transfer mode. A connection-oriented network technology that uses small, fixed-size packets (called cells) to carry data.


ATMARP: Address Resolution Protocol as enhanced for ATM networks. ATM Forum: A key ATM standards-setting body.


authentication: Security protocol by which two suspicious parties prove to each other that they are who they claim to be.


autonomous system (AS): A group of networks and routers, subject to a common authority and using the same intradomain routing protocol.


bandwidth: A measure of the capacity of a link or connection, usually given in units of bits per second.


Bellman-Ford: A name for the distance-vector routing algorithm, from the names of the inventors.


BER: Basic encoding rules. Rules for encoding data types defined by ASN.1. best-effort delivery: The service model of the current Internet architecture. Delivery of a message is attempted but is not guaranteed.


BGP: Border Gateway Protocol. An interdomain routing protocol by which autonomous systems exchange reachability information. The most recent version is BGP-4.


BISYNC: Binary Synchronous Communication. A byte-oriented link-level protocol developed in the late 1960s by IBM.


bit stuffing: A technique used to distinguish control sequences and data on the bit level. Used by the HDLC protocol.


block: An OS term used to describe a situation in which a process suspends execution while awaiting some event, such as a change in the state of a semaphore.


Bluetooth: A short-range wireless standard used to connect computers, mobile phones, and peripheral devices, among other things.


bridge: A device that forwards link-level frames from one physical network to another, sometimes called a LAN switch. Contrast with repeater and router.

broadcast: A method of delivering a packet to every host on a particular network or internet. May be implemented in hardware (e.g., Ethernet) or software (e.g., IP broadcast).


CA: Certification authority (also known as certificate authority). An entity that signs security certificates, thereby promising that the public key contained in the certificate belongs to the entity named in the certificate.


CBC: Cipher block chaining. A cryptographic mode in which each plaintext block is XORed with the previous block of ciphertext before encryption.


CBR: Constant bit rate. A class of service in ATM that guarantees transmission of data at a constant bit rate, thus emulating a dedicated transmission link. Contrast with ABR, UBR, and VBR.


CCITT: The now defunct Comité Consultif International de Telegraphique et Telephonique, a unit of the International Telecommunications Union (ITU) of the United Nations.


Now replaced by ITU-T.

CDMA: Code Division Multiple Access, a form of multiplexing used in wireless networks.


CDN: Content distribution network. A collection of surrogate web servers, distributed across the Internet, that respond to web HTTP requests in place of the server. The goal of widely distributing the surrogate servers is to have a surrogate close to the client, making it possible to respond to requests more quickly.


cell: A 53-byte ATM packet, capable of carrying up to 48 bytes of data.

certificate: A document digitally signed by one entity that contains the name and public key of another entity. Used to distribute public keys. Also see CA.


channel: A generic communication term used in this book to denote a logical processto-process connection.


checksum: Typically a ones complement sum over some or all of the bytes of a packet, computed and appended to the packet by the sender. The receiver recomputes the checksum and compares it to the one carried in the message. Checksums are used to detect


errors in a packet and may also be used to verify that the packet has been delivered to the correct host. The term checksum is also sometimes (imprecisely) used to refer generically to error-detecting codes.


chipping code: Random sequence of bits that is XORed with the data stream to implement the direct sequence technique of spread spectrum.


CIDR: Classless interdomain routing. A method of aggregating routes that treats a block of contiguous Class C IP addresses as a single network.


circuit switching: A general strategy for switching data through a network. It involves establishing a dedicated path (circuit) between the source and destination. Contrast with packet switching.

client: The requester of a service in a distributed system.


CLNP: Connectionless Network Protocol. The ISO counterpart to the Internet’s IP. clock recovery: The process of deriving a valid clock from a serially transmitted digital signal.


concurrent logical channels: Multiplexing several stop-and-wait logical channels onto a single point-to-point link. No delivery order is enforced. This mechanism was used by the IMP-IMP protocol of the ARPANET.


congestion: A state resulting from too many packets contending for limited resources (e.g., link bandwidth and buffer space on routers or switches), which may force the router (switch) to discard packets.

congestion control: Any network resource management strategy that has, as its goal, the alleviation or avoidance of congestion. A congestion-control mechanism may be implemented on the routers (switches) inside the network, by the hosts at the edges of the


network, or by a combination of both.

connection: In general, a channel that must be established prior to use (e.g., by the transmission of some setup information). For example, TCP provides a connection abstraction


that offers reliable, ordered delivery of a byte stream. Connection-oriented networks, such as ATM, are often said to provide a virtual circuit abstraction.


connectionless protocol: A protocol in which data may be sent without any advance setup. IP is an example of such a protocol.


context switch: An operation in which an operating system suspends the execution of one process and begins the execution of another. A context switch involves saving the state of the former process (e.g., the contents of all registers) and loading the state of the latter process.


controlled load: One of the service classes available in the Internet’s Integrated Services




CRC: Cyclic redundancy check. An error-detecting code computed over the bytes composing a packet and then appended to the packet by the network hardware (e.g., Ethernet


adaptor). CRC provides stronger error detection than a simple checksum.

crossbar switch: A simple switch design in which every input is directly connected to every output and the output port is responsible for resolving contention.


CSMA/CD: Carrier Sense Multiple Access with Collision Detect. CSMA/CD is a functionality of network hardware. “Carrier sense multiple access” means that multiple stations

can listen to the link and detect when it is in use or idle; “collision detect” indicates that if two or more stations are transmitting on the link simultaneously, they


will detect the collision of their signals. Ethernet is the best-known technology that uses CSMA/CD.


cut-through: A form of switching or forwarding in which a packet starts to be transferred to an output before it has been completely received by the switching node, thus reducing latency through the node.


datagram: The basic transmission unit in the Internet architecture. A datagram contains all of the information needed to deliver it to its destination, analogous to a letter in the U.S. postal system. Datagram networks are connectionless.


DCE: Distributed Computing Environment. An RPC-based suite of protocols and standards that support distributed computing. Defined by OSF.


DDCMP: DigitalData CommunicationMessage Protocol. A byte-oriented link-level protocol used in Digital Equipment Corporation’s DECNET.


DDoS: Distributed denial of service. A DoS attack in which the attack originates at a set of nodes. Each attacking node may put only a marginal load on the target machine, but the aggregate load from all the attacking nodes swamps the target machine.


DECbit: A congestion-control scheme in which routers notify the endpoints of imminent congestion by setting a bit in the header of routed packets. The endpoints decrease their sending rates when a certain percentage of received packets have the bit set.


decryption: The act of reversing an encryption process to recover the data from an encrypted message.

delay bandwidth product: The product of a network’s RTT and bandwidth. Gives a measure of how much data can be in transit on the network.


demultiplexing: Using information contained in a packet header to direct it upward through a protocol stack. For example, IP uses the ProtNum field in the IP header to


decide which higher protocol (i.e., TCP, UDP) a packet belongs to, and TCP uses the port number to demultiplex a TCP packet to the correct application process. Contrast with multiplexing.


demultiplexing key: A field in a packet header that enables demultiplexing to take place (e.g., the ProtNum field of IP).


dense mode multicast: PIM mode used when most routers or hosts need to receive multicast packets.


DES: Data Encryption Standard. An algorithm for data encryption based on a 64-bit secret key.


DHCP: Dynamic Host Configuration Protocol. A protocol used by a host as it boots or when it is connected to a network, to learn various network information, such as its IP address.


DHT: Distributed hash table. A technique by which a message is routed toward a machine that supports a particular object, based on the object’s name. The object is hashed to a


unique identifier, with each intermediate node along the route forwarding the message to a node that is able to interpret a larger prefix of this ID. DHTs are often used in peer-to-peer networks.


Differentiated Services: A new architecture for providing better than best-effort service on the Internet. It has been proposed as an alternative to Integrated Services.


direct sequence: A spread spectrum technique that involves XORing the data stream with a random bit sequence known as a chipping code.


distance vector: A lowest-cost-path algorithm used in routing. Each node advertises reachability information and associated costs to its immediate neighbors, and uses the updates it receives to construct its forwarding table. The routing protocol RIP uses a distance-vector algorithm. Contrast with link state.


DMA: Direct memory access. An approach to connecting hosts to I/O devices, in which the device directly reads data from and writes data to the host’s memory. Also see PIO.

DNA/DECNET: Digital Network Architecture. An OSI-based architecture that supports a connectionless network model and a connection-oriented transport protocol.


DNS: Domain name system. The distributed naming system of the Internet, used to resolve host names (e.g., cicada.cs.princeton.edu) into IP addresses (e.g., The DNS is implemented by a hierarchy of name servers.


domain: Can refer either to a context in the hierarchical DNS namespace (e.g., the “edu” domain) or to a region of the Internet that is treated as a single entity for the purpose of hierarchical routing. The latter is equivalent to autonomous system.


DoS: Denial of service. A situation in which an attacking node floods a target node with so much work (so many packets) that it effectively keeps legitimate users from accessing the node, hence, they are denied service.


DS3: A 44.7-Mbps transmission link service offered by the phone company. Also called T3.


DSL: Digital subscriber line. A family of standards for transmitting data over twisted pair telephone lines at multimegabit-per-second speeds.

duplicate ACK: A retransmission of a TCP acknowledgment. The duplicate ACK does not acknowledge any new data. The receipt of multiple duplicate ACKs triggers the TCP fast retransmit mechanism.


DVMRP: Distance Vector Multicast Routing Protocol. Multicast routing protocol originally used in the MBone.

DWDM: Dense wavelength division multiplexing. Multiplexing multiple light waves


(colors) onto a single physical fiber. The technique is “dense” in the sense that a large number of optical wavelengths can be supported.


ECN: Explicit congestion notification. A technique by which routers inform end hosts about congestion by setting a flag in packets they are forwarding. Used in conjunction with active queue management algorithms like RED.


EF: Expedited forwarding. One of the per-hop behaviors proposed for Differentiated Services.


EGP: Exterior Gateway Protocol. An early interdomain routing protocol of the Internet, which was used by exterior gateways (routers) of autonomous systems to exchange routing information with other ASs. Replaced by BGP.


encapsulation: The operation, performed by a lower-level protocol, of attaching a protocol-specific header and/or trailer to a message passed down by a higher-level protocol. As a message travels down the protocol stack, it gathers a sequence of headers, of


which the outermost corresponds to the protocol at the bottom of the stack. encryption: The act of applying a transforming function to data, with the intention that only the receiver of the data will be able to read it (after applying the inverse function,


decryption). Encryption generally depends on either a secret shared by the sender and receiver or on a public/private key pair.


Ethernet: A popular local area network technology that uses CSMA/CD and has a bandwidth of 10 Mbps. An Ethernet itself is just a passive wire; all aspects of Ethernet transmission

are completely implemented by the host adaptors.

exponential backoff: A retransmission strategy that doubles the timeout value each time a packet is retransmitted.


exposed node problem: Situation that occurs on a wireless network where two nodes receive signals from a common source, but each is able to reach other nodes that do not receive this signal.


extended LAN: A collection of LANs connected by bridges.

fabric: The part of a switch that actually does the switching, that is, moves packets from input to output. Contrast with port.


fair queuing (FQ): A round-robin-based queuing algorithm that prevents a badly behaved process from capturing an arbitrarily large portion of the network capacity.


fast retransmit: A strategy used by TCP that attempts to avoid timeouts in the presence of lost packets. TCP retransmits a segment after receiving three consecutive duplicate ACKs, acknowledging the data up to (but not including) that segment.


FDDI: Fiber Distributed Data Interface. A token ring networking technology designed to run over optical fiber.


FEC: 1 Forward error correction. A general strategy for recovering from bit errors introduced into data packets without having to retransmit the packet. Redundant


information is included with each packet that can be used by the receiver to determine which bits in a packet are incorrect. Contrast with ARQ.

2 Forwarding equivalence class. A set of packets that are to receive the same forwarding treatment at a router. MPLS labels are normally associated with FECs.


Fibre Channel: A bidirectional link protocol commonly used to connect computers, peripherals, and storage devices. Originally had a bandwidth of 100 MBps but since enhanced


to GBps speeds.

firewall: A router that has been configured to filter (not forward) packets from certain sources. Used to enforce a security policy.


flow control: A mechanism by which the receiver of data throttles the transmission rate of the sender, so that data will not arrive too quickly to be processed. Contrast with

congestion control.


flowspec: Specification of a flow’s bandwidth and delay requirements presented to the network to establish a reservation. Used with RSVP.


forwarding: The operation performed by a router on every packet: receiving it on an input, deciding what output to send it to, and sending it there.


forwarding table: The table maintained in a router that lets it make decisions on how to forward packets. The process of building up the forwarding table is called routing, and thus the forwarding table is sometimes called a routing table. In some implementations, the routing and forwarding tables are separate data structures.


fragmentation/reassembly: A method for transmission of messages larger than the network’s


MTU. Messages are fragmented into small pieces by the sender and reassembled by the receiver.


frame: Another name for a packet, typically used in reference to packets sent over a single link rather than a whole network. An important problem is how the receiver detects the beginning and ending of a frame, a problem known as framing.


Frame Relay: A connection-oriented public packet-switched service offered by the phone company.


frequency hopping: A spread spectrum technique that involves transmitting data over a random sequence of frequencies.


FTP: File Transfer Protocol. The standard protocol of the Internet architecture for transferring files between hosts. Built on top of TCP.


GMPLS: Generalized MPLS. Allows IP to run natively over optically-switched networks. GPRS: General Packet Radio Service. A packet transmission service provided by cellular wireless networks.


GSM: Global System for Mobile communication. Digital cellular phone system being deployed throughout the world (less so in the United States and Canada). Similar to PCS, which is being deployed throughout the United States and Canada.


gopher: An Internet information service.


H.323: Session control protocol often used for Internet telephony.


handle: In programming, an identifier or pointer that is used to access an object. hardware address: The link-level address used to identify the host adaptor on the local network.


HDLC: High-Level Data Link Control protocol. An ISO-standard link-level protocol. It uses bit stuffing to solve the framing problem.


hidden node problem: Situation that occurs on a wireless network where two nodes are sending to a common destination, but are unaware that the other exists.

hierarchical routing: A multilevel routing scheme that uses the hierarchical structure of the address space as the basis for making forwarding decisions. For example, packets might first be routed to a destination network and then to a specific host on that network.


HiPPI: High Performance Parallel Interface. An ANSI-standard network technology capable of Gbps transmission rates, typically used to connect supercomputers to peripheral


devices. Used in same way as Fibre Channel.

host: A computer attached to one or more networks that supports users and runs application programs.


HTML: HyperText Markup Language. A language used to construct World Wide Web pages.


HTTP: HyperText Transport Protocol. An application-level protocol based on a request/ reply paradigm and used in the World Wide Web. HTTP uses TCP connections


to transfer data.

IAB: Internet Architecture Board. The main body that oversees the development of the Internet architecture.


IBGP: Interior BGP. The protocol used to exchange interdomain routing information among routers in the same domain.


ICMP: Internet ControlMessage Protocol. This protocol is an integral part of IP. It allows a router or destination host to communicate with the source, typically to report an error


in IP datagram processing.

IEEE: Institute for Electrical and Electronics Engineers. A professional society for engineers that also defines network standards, including the 802 series of LAN standards.


IETF: Internet Engineering Task Force. The body responsible for the specification of standards and protocols related to the Internet.


IMAP: InternetMessage Access Protocol. An application layer protocol that allows a user to retrieve her email from a mail server.


IMP-IMP: A byte-oriented link-level protocol used in the original ARPANET.

Integrated Services: Usually taken to mean a packet-switched network that can effectively support both conventional computer data and real-time audio and video. Also, a


name given to a proposed Internet service model that was designed to supplement the current best-effort service model.


integrity: In the context of network security, a service that ensures that a received message is the same one that was sent.


interdomain routing: The process of exchanging routing among different routing domains. BGP is an example of an interdomain protocol.


internet: A collection of (possibly heterogeneous) packet-switching networks interconnected by routers. Also called an internetwork.


Internet: The global internet based on the Internet (TCP/IP) architecture, connecting millions of hosts worldwide.


interoperability: The ability of heterogeneous hardware and multivendor software to communicate by correctly exchanging messages.


interrupt: An event (typically generated by a hardware device) that tells the operating system to stop its current activity and take some action. For example, an interrupt is used to notify the OS that a packet has arrived from the network.


intradomain routing: The exchange of routing information within a single domain or autonomous system. RIP and OSPF are example intradomain protocols.

IP: Internet Protocol (also known as IPv4). A protocol that provides a connectionless, best-effort delivery service of datagrams across the Internet.

IPng: Internet Protocol—Next Generation (also known as IPv6). Proposed version of IP that provides a larger, more hierarchical address space and other new features.


IPSEC: IP Security. An architecture for authentication, privacy, and message integrity, among other security services to the Internet architecture.


IRTF: Internet Research Task Force. A sibling body to the IETF, responsible for charting direction in research and development for the Internet.


IS-IS: A link-state routing protocol, similar to OSPF.


ISDN: Integrated Services Digital Network. A digital communication service offered by telephone carriers and standardized by ITU-T. ISDN combines voice connection and digital data services in a single physical medium.


ISO: International Standards Organization. The international body that drafted the seven-layer OSI architecture and a suite of protocols that has not enjoyed commercial success.


ITU-T: A subcommittee of the International Telecommunications Union, a global body


that drafts technical standards for all areas of international analog and digital communication. ITU-T deals with standards for telecommunications, notably ATM.


jitter: Variation in network latency. Large jitter has a negative impact on the quality of video and audio applications.


JPEG: Joint Photographic Experts Group. Typically used to refer to a widely used algorithm for compressing still images that was developed by the JPEG.


Kerberos: A TCP/IP-based authentication system developed atMIT, in which two hosts use a trusted third party to authenticate each other.

key distribution: Mechanism by which users learn each others’ public keys through the exchange of digitally signed certificates.


LAN: Local area network. A network based on any physical network technology that is designed to span distances of up to a few thousand meters (e.g., Ethernet or FDDI). Contrast with SAN, MAN, and WAN.

Study Material, Lecturing Notes, Assignment, Reference, Wiki description explanation, brief detail
Computer Networks : Important Glossary in Computer Networks |

Privacy Policy, Terms and Conditions, DMCA Policy and Compliant

Copyright © 2018-2023 BrainKart.com; All Rights Reserved. Developed by Therithal info, Chennai.