1、_ 1rJ-T RECMN*E.457 96 m 4862591 Ob37352 727 D INTERNATIONAL TELECOMMUNICATION UNION ITU-T TELECOMMUNICATION STANDARDIZATION SECTOR OF ITU E.457 (OU96) TELEPHONE NETWORK AND ISDN QUALITY OF SERVICE, NETWORK MANAGEMENT AND TRAFFIC ENGINEERING FACSIMILE MEASUREMENT METHODOLOGIES ITU-T Recommendation E
2、.457 (Previously CCITT Recommendation“) ITU-T RECMN*E.457 76 4862571 Ob17353 bb3 = FOREWORD The -T (Telecommunication Standardization Sector) is a permanent organ of the International Telecommunication Union (rrv). The ITU-T is responsible for studying technical, operating and ariff questions and is
3、suing Recommen- dations on them with a view to standardizing telecommunications on a worldwide basis. The World Telecommunication Standardization Conference (WTSC), which meets every four years, establishes the topics for study by the ITU-T Study Groups which, in their turn, produce Recommendations
4、on these topics. The approval of Recommendations by the Members of the IT-T is covered by the procedure laid down in WTSC Resolution No. 1 (Helsinki, March 1-12, 1993). ITU-T Recommendation E.457 was prepared by -T Study Group 2 (1993-1996) and was approved under the WTSC Resolution No. 1 procedure
5、on the 19th of February 1996. NOTE In this Recommendation, the expression “Administration” is used for conciseness to indicate both a telecommunication administration and a recognized operating agency. O 1996 All rights reserved. No part of this publication may be reproduced or utilized in any form
6、or by any means, electronic or mechanical, including photocopying and microfilm, without permission in writing from the ITU. ITU-T RECHNaE.457 96 W 4862591 061735 5TT = CONTENTS 1 References 2 Definitions 3 Abbreviations . 4 Introduction 5 Measurement methodologies 5.1 Test calls . 5.2 Non-intrusive
7、 measurements 5.3 Mixed mode measurements 5.4 Terminal based measurements 6 Test planning 7 statistical aspects Annex A - Estimation of sample sizes for facsimile performance parameters . Page 1 1 1 1 2 2 3 4 4 4 5 5 Recommendation E.457 (02196) i ITU-T RECMN*E.457 7b 4b2591 Ob17355 436 W SUMMARY Th
8、is Recommendation discusses various facsimile measurement methodologies focusing on G3 facsimile as defined by Recommendations T.4, T.6. T.30 and used on PSTN as defined by Recommendations F.180 and F.182. The Recommendation evaluates the advantages and disadvantages of intrusive (test-call based),
9、non-intrusive, mixed-mode and terminai based measurements. Test planning aspects including planning approach, measurement topology, test location points, and some statistical aspects are also discussed ii Recommendation E.457 (02/96) ITU-T RECMNaE-457 96 Y862591 0617356 372 m Recommendation E.457 FA
10、CSIMILE MEASUREMENT METHODOLOGIES (Geneva, 1996) 1 References The following Recommendations contain material relevant to this Recommendation. - ITU-T Recommendation T.4 (1993), Standardization of Group 3 facsimile apparatus for document transmission. CCT Recommendation T.6 (1988), Facsimile coding s
11、chemes and coding controlfunctions for Group 4 facsimile apparatus. ITU-T Recommendation T.30 (1993), Procedures for document facsimile transmission in the general switched telephone network. ITU-T Recommendation F. 182 (1993), Operational provisions for the intemutional public facsimile service bet
12、ween subscribers stations with Group 3 facsimile machines (telefa 3). ITU-T Recommendation E.450 (1993), Facsimile quality of service on PSTN - General aspects. IT-T Recommendation E.45 1 (1993), Facsimile call cut-offperformance. ITU-T Recommendation E.452 (1993). Facsimile modem speed reductions a
13、nd transaction time. ITU-T Recommendation E.453 (1994), Facsimile image quality as corrupted by transmission-induced scan line errors. II7J-T Recommendation E.456 (1994), Test transaction for facsimile transmission performance. - - - - - - - - 2 Definitions For the purposes of this Recommendation th
14、e definitions given in Recommendations E.450, E.45 1, E.452 and E.453 will be applicable. 3 Abbreviations A glossary is presented in Recommendation E.450. 4 Introduction The intent of this Recommendation is to provide an overview of facsimile measurement methodologies, their advantages and disadvant
15、ages, test planning and statistical aspects of measurements. The measurement of facsimile service quality of networks can be accomplished by a variety of techniques including: 4.1 developed in the future. 4.2 measurement equipment meeting certain minimum requirements will be defined in the future. 4
16、.3 monitoring for trouble shooting. Intrusive test-call-based measurements using test machines meeting certain minimum requirements will be Non-intrusive measurements based on live customer traffk monitored at a network interface using Mixed mode measurements involving test calls utilizing non-intru
17、sive (protocol and analogue signal) 4.4 In the following, we discuss these measurement methodologies in some detail. Terminal based measurements obtained from terminal reports meeting certain technical requirements. Recommendation E.457 (02196) 1 ITU-T RECMN*E.457 76 48b2591 Ob17357 207 5 Measuremen
18、t methodologies 5.1 Test calls Measurements using test-call-based methodology are typically carried out over an end-to-end path. The general test topology is shown in Figure 1. Usually, the test calls interface the network at a point similar to that of a typical subscriber. It is recommended that th
19、e test calls follow the prescriptions listed below: 5.1.1 Test calls use test transactions defined in Recommendation E.456. 5.1.2 5.1.3 5.1.4 Performance parameters be measured as specified in Recommendations E.41, E. Test machines satisfy the minimum requirements. (To be developed.) Statistical req
20、uirements are met (see clause 7). 52, and E.J 53. TS CO TW4930-95/dO1 International netwark CO National nehnrork 2W 2wire 4W 4wire CO Centraioffice TS Tollswitch ISC International switching cmtre FIGURE 1E.457 Example of intrusive measurement topdogy 5.1.5 Remarks Test calls provide a means for end-
21、to-end performance characterization of networks in a controlled fashion. Test calls can be used for the comparison of performance of specific network services, network technologies, customer premises equipment, etc. Test calls are one of the faster methods of trouble-shooting a network problem or a
22、networklterminal interaction problem. Test calls are costly in the sense that they use up network resources for testing purposes. Test calls provide only a limited view of the geographical diversity since it is usually not possible to provide and maintain a large number of test access points. From p
23、ractical considerations, the test access points are located in major communication centres which presumably have a higher quality network performance than, say, the outlying areas. Test calls tend to produce performance results that are in general better than that observed in non- intrusive measurem
24、ents because the impact of customer CPE malfunctioning and incompatibilities, poor quality access and egress networks, etc. is usually avoided. 2 Recommendation E.457 (02/96) ITU-T RECMN*E*457 96 = 4862573 Ob17358 I145 g) When conducting network tests, the prevailing conditions should be known to th
25、e maximum possible extent. For example, if the comparative performance of different media is being assessed, the network should be in its nominal state. If abnormal conditions occur (e.g. network under restoration) during tests, this data should be analysed separately. 5.2 Non-intrusive measurements
26、 Usually, non-intrusive measurement systems monitor the performance of facsimile calls at a network concentration point such as a switch (Figure 2). However, it is possible to monitor facsimile performance non-intrusively at any point in the connection with the appropriate interface. In general, non
27、-intrusive monitoring provides an end-to-end view of facsimile performance regardless of the measurement point. Some network equipment with FAX DEMOD and REMOD may be provisioned so that the G3 fax protocol messages are intentionally altered (e.g. limit the maximum speed) and the characteristics of
28、facsimile transactions are changed as a result. In this case, the measurements should be analysed taking into account the effects of such equipment. 5.2.1 will be defined in the future. It is recommended that non-intrusive measurements be conducted using facsimile performance parameters that Non-int
29、ruskm fax rneasuement system 2W 2wire 4W 4wire CO Ceniraloffice TS Tdlswitch ISC International switching centre FIGURE uE.457 Example of non-intnisive measurement tapdogy 5.2.2 Remarks a) b) c) Non-intrusive measurements provide a means for characterizing the end-to-end service quality of networks,
30、including a variety of CPE interactions, customer settings and malfunctions. Depending on the capabilities of the equipment, non-intrusive measurements could also provide a comparison of the performance of specific network services, network technologies, CPE, etc. Such measurements are usually able
31、to gather data quickly and in large volumes once the network arrangements for the data collection are completed. The ability to capture data is a function of the fax traffk volume and the number of channels that can be monitored simultaneously. Measurement costs are limited to costs for data collect
32、ion and analysis, and network resources are not tied up for the purposes of testing. Measurements can provide a general view of tail network conditions depending on where the measurement point is located. If it is possible to associate the per call performance measurements with calling and called nu
33、mbers, it is possible to obtain a very detailed view of tail network performances. d) e) Recommendation E.457 (0296) 3 ITU-T RECMN*E-457 96 W 4662591 Ob17359 O81 a f) Measurements produce results that are in general worse than those observed in test-call based measurements because of the CPE malfunc
34、tioning and incompatibilities, poor quality access and egress networks, etc. g) While non-intrusive measurements offer certain advantages, they do not permit the level of control that can be achieved using the test call methodology. For example, when sending test calls all transactions can use test
35、transactions defined in Recommendaion E.456 providing a very controlled test stimulus. Such control is not possible for non-intrusive measurements. h) When non-intrusive monitoring is conducted at a network concentration point, image quality parameters defined in Recommendation E.453 cannot be evalu
36、ated directly. However, ECM based metrics could be characterized which could serve as a surrogate for the image quality parameters defined in Recommenda- tion E.453. Additionally, analysis of RTN and RTP signals can be used to give a general picture of image degradation due to transmission errors. 5
37、.3 Mixed mode measurements Mixed mode measurements may be used for specialized purposes such as trouble-shooting. In this case, test calls are launched from a test apparatus which could include a facsimile machine that is having problems of some kind. Non-inrusive equipment could be as simple as a T
38、.30 protocol analyser or could be a more complex device capable of performing both T.30 and analogue signal characteristic measurements. There are no specific requirements that pertain to these types of trouble-shooting measurements except that the measurements be technically sound. 5.4 Terminal bas
39、ed measurements Facsimile terminals are capable of producing a variety of terminal reports indicating the success/failure of calls, call duration, some protocol information for failed calls, etc. The number and selection of items varies by manufacturer and model. These reports are more useful than h
40、aving no information at all. However, there are no IT standards for how this information is gathered by the machine. Therefore, caution should be exercised in interpreting the data from terminal reports. This is especially true for transaction times reported by terminals where considerable variation
41、s exist in the calculation of this parameter. 6 Test planning Planning of a facsimile performance study should include careful consideration of the goals of the study, primary factors that are planned to be studied, secondary factors, interactions, statistical aspects, etc. a) The tests must be plan
42、ned such that the results for the primary factors can be obtained without confounding those factors. In formulating the test plan, interactions that are likely to arise must be identified beforehand and the interpretations of the test results must include the possible effects of these interactions.
43、For example, when comparing the facsimile performance on two types of transport media on the international segment, it is important to keep the tail network conditions as similar. b) Secondary factors such as time of day can often be important in network tests because the load on the networks, routi
44、ng, etc. can be different during different periods. The sample plan should include these factors if it is likely that the results and their interpretation would be affected substantially. c) It is often useful to conduct a pilot study with a smaller sample size prior to conducting the full study. Th
45、e pilot study can often provide guidance on what factors are important, what are the likely problems to be encountered or what the statistical considerations are likely to be, etc. d) Test machines used for intrusive testing should be carefully selected so that they have no known incompatibility pro
46、blems, are representative of the facsimile terminal base, etc. They should meet the requirements which will be specified in the future. 4 Recommendation E.457 (0296) ITU-T RECMN*E.457 ?76 4862593 ObL73b0 BT3 7 Statistical aspects Results of facsimile performance tests must be statistically valid and
47、 special care must be exercised to ensure that a sufficient number of samples are collected and the correct statistical analysis is employed. Depending on the questions to be studied, the test hypothesis must be formulated and the sample size estimates developed. There are a variety of texts that ar
48、e available for this purpose and some of them are listed in the references. As an example, the sample size for estimating facsimile call cut-off ratios at a given level of accuracy is presented in Annex A. Additional details on statistical questions are proposed to be included in the facsimile QOS H
49、andbook. Annex A Estimation of sample sizes for facsimile performance parameters (This annex forms an integral part of this Recommendation) The E-Series Recommendations ?E.45 1 : Facsimile call cut-off performance?, ?E.452: Facsimile modem speed reductions and transaction time? and ?E.453: Facsimile image quality as corrupted by transmission-induced scan line errors? define and formulate a number of dichotomous parameters expressed in terms of percentages. These parameters are frequently estimated by conducting field studies. To ensure the statistical validity
