oo0wMMOWT oo0wMm0WT
`NETWORK-1 N1-2008
`
`
`
`
`

`
`OXFORD
`UNIVERSITY PRESS
`Great Clarendon Street, Oxford OX2 6DP
`
`Oxford University Press is a department of the University of Oxford.
`It furthers the University’s objective of excellence in research, scholarship,
`and education by publishing worldwide in
`Oxford New York
`Auckland Bangkok Buenos Aires Cape Town Chennai
`Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata
`Kua1aLumpur Madrid Melbourne Mexico City Mumbai Nairobi
`Szio Paulo Shanghai Singapore Taipei Tokyo Toronto
`
`Oxford is a registered trade mark of Oxford University Press
`in the UK and in certain other countries
`
`© Market House Books Ltd. 1983. 1986, 1990. 1996, 2004
`
`The moral rights of the author have been asserted
`Database right Oxford University Press (maker)
`First published 1983
`Second edition 1986
`Third edition 1990
`Fourth edition 1996
`Fifth edition 2004
`
`All rights reserved. No part of this publication may be reproduced,
`stored in a retrieval system. or transmitted in any form or by any means,
`without the prior permission in writing of Oxford University Press.
`or as expressly permitted by law. or under terms agreed with the appropriate
`reprographics rights organization. Enquiries concerning reproduction
`outside the scope of the above should be sent to the Rights Department.
`Oxford University Press, at the address above
`You must not circulate this book in any other binding or cover
`and you must impose this same condition on any acquirer
`British Library Cataloguing in Publication Data
`Data available
`
`Library of Congress Cataloging in Publication Data
`Data available
`
`ISBN 0-19-860877~2
`
`1 T
`
`ypeset in Swift by Market House Books Ltd.
`Printed in Great Britain by Clays Ltd, St Ives pic
`
`
`
`
`
`

`
`
`
`data processing
`
`data model An abstract model of some
`real-world situation or domain of interest
`about which information is to be held in a
`*database and which the *logical schema
`for that database encodes. The term data
`model (or data modeling method) is also
`used for a set of logical abstractions em-
`ployed in constructing such a model. See
`also RELATIONAL MODEL, HIERARCHICAL
`DATA MODEL, CODASYL NETWORK MODEL.
`
`Data Module The name used by IBM to
`refer to their removable, hermetically
`sealed disk pack, incorporating the
`read/write heads and carriage assembly,
`that was used with the 3340 *Winchester
`technology disk drive. Current data-
`processing systems use fixed disk storage;
`however, the term data module was once
`in general use and was interchangeable
`with the terms disk pack and storage mod-
`ule.
`
`data name In data-processing languages,
`a symbolic name chosen by the program-
`mer to identify a data object. See also VARI-
`ABLE.
`
`data network A communication net-
`work that is devoted to carrying computer
`information, as opposed to voice, video,
`etc. It consists of a number of nodes, or
`stations, connected by various communi-
`cation channels.
`
`data path Another namefor data bus, al-
`though often used in a wider context to
`mean any logical or physical connection
`between a source and destination of digi-
`tal or analog information.
`data preparation An obsolete term for
`the process of converting data into a ma-
`chine-readable form so that it can be en-
`tered into a system via an available input
`device. There is no interaction with the
`system in the course of preparation. The
`process has been superseded by direct
`*data entry systems and *data capture.
`data processing (DP) A term used pre-
`dominantly in the context of industrial,
`business, governmental, and other organi-
`zations: within that context it refers (a) to
`a class of computer applications, (b) to a
`function within the organization.
`While it is hard to generalize, data-
`processing applications may be character-
`ized as those that store and process large
`
`
`
`134
`
`135
`
`hat any datagram
`l before reaching its
`in which datagrams
`aurce is not neces-
`lelivery. In some
`iossibility that a
`.cated and delivered
`3 than once. It is
`e application to
`ising from data-
`n.
`
`rarchical structure
`3) a record at level 1'
`non to a set of
`cl (b) starting from
`, it is possible to ac-
`vel records. Any
`‘ one set of lower-
`only be a member
`hierarchy may
`rarchical relation-
`zr view provided by
`iurpose or activity.
`The facility to mod-
`j*logical or *storage
`‘quent requirement
`or programs inter-
`se nor any need to
`- data independence
`ation for the devel-
`anagernent soft-
`rm and different
`rent levels of data
`'ticularly important
`[SE8 that are re-
`: with user needs.
`ndependence fre-
`the need for effi-
`ng and usually
`promise in terms of
`as used. Logical data
`he facility to change
`l thus evolve the
`e; physical data inde-
`acility to change the
`us modify and im-
`
`:ance to alteration
`ta stored in a com-
`that denotes only
`‘ alteration of data.
`eliability of data
`edia, in terms of the
`errors (see ERRoR
`detected error rate
`
`perceived by the host system may be
`worse than that arising at the magnetic
`disk or tape if undetected errors can arise,
`eg. from the eifect of noise on connecting
`cables where the interface concerned has
`insufficient error detection capability.
`From a system point of view_ the unde-
`tected error rate of a peripheral may be in-
`adequate: the system can improve on it by
`making additional provision for checking
`in software.
`
`data item The *representation of any
`value that can be used alone or as a com-
`ponent of a *data structure.
`data link A physical connection between
`two or more devices (called nodes or sta-
`tions) by a communication channel that ap-
`pears “wirelike”, i.e. bits arrive in the
`order sent. Coaxial cables, telephone lines,
`optical fibers, lasers, and even satellite
`channels can be data links. Data links are
`assumed to be susceptible to noise (i.e.
`have error properties) and have finite data
`rate and nonzero propagation delay.
`data link control protocol A commu-
`nication *protocol that converts noisy
`(error—prone) *data links into communica-
`tion channels free of transmission errors.
`Data is broken into *frames, each of
`which is protected by *checksum. Frames
`are retransmitted as many times as
`needed to accomplish correct transmis-
`sion. A data link control protocol must
`prevent data loss caused by mismatched
`sending/receiving capacities. A *flow con-
`trol procedure, usually a simple sliding
`*window mechanism, provides this func-
`tion. Data link control protocols must pro-
`vide *transparent data transfer. *Bit
`stuffing or byte stufling strategies are used
`to mask control patterns that occur in the
`text being transmitted. Control frames are
`used to start/stop logical connections over
`links. *Addressing may be provided to sup-
`port several *virtual connections on the
`same physical link.
`data link layer of network protocol
`function. See SEVEN-LAYER REEERENCE
`MODEL.
`
`data logging A procedure that involves
`recording all data and interactions that
`pass through a particular point in a sys-
`tem. The point chosen is usually part of a
`communication loop or a data path to or
`
`from a device such as a keyboard and dis-
`play on which data is transitory. lf a sys-
`tem failure or an unexpected result occurs
`it is possible to reconstruct the situation
`that existed. Such logs are not generally
`archived and can be overwritten once the
`associated job has been completed. See also
`DATA CAPTURE.
`
`data management A term normally
`used to refer to systems that offer users an
`interface that screens them from the ma-
`jority of the details of the physical han-
`dling of the files, leaving them free to
`concentrate on the logical properties of
`the data.
`
`data management system A class of
`software systems that includes *database
`management systems and *file manage-
`ment systems.
`
`data manipulation, data manipula-
`tion language See DATABASE LANGUAGE.
`data mark See ADDRESS MARK.
`
`data matrix A rectangular array of data
`variables, which may be numerical,
`classificatoiy, or alphanumeric. The data
`matrix forms the input structure upon
`which statistical procedures for *regres-
`sion analysis, *analysis of variance, *mul-
`tivariate analysis, *cluster analysis, or
`survey analysis will operate.
`data medium A material having defined
`properties, including a physical variable
`that can be used to represent data. The
`defined properties ensure that the
`medium is compatible with devices that
`can record or read data on the medium.
`Examples of data media are *magnetic
`tape, *Inagnetic disks, and *optical disks,
`and also paper used for printer output.
`data mining The nontrivial explication
`or extraction of information from data, in
`which the information is implicit and pre-
`viously unknown; an example is identifica-
`tion of the pattern of use of a credit card
`to detect possible fraud. The data is nor-
`mally accessed from one or more data-
`bases, so the technique is also known as
`knowledge discovery in databases (KDD). lt in-
`volves a number of different methods
`from artificial intelligence such as neural
`networks and machine induction, to-
`gether with statistical methods such as
`cluster analysis and data summarization.
`
`y