1、1Tid-T RECMN8E.493 96 m 4862593 Ob37363 502 INTERNATIONAL TELECOMMUNICATION UNION ITU-T TELECOMMUNICATION STANDARDIZATION SECTOR OF ITU E.493 (02/96) TELEPHONE NETWORK AND ISDN QUALITY OF SERVICE, NETWORK MANAGEMENT AND TRAFFIC ENGINEERING GRADE OF SERVICE (GOS) MONITORING ITU-T Recommendation E.493
2、 (Previously “CCITT Recommendation”) ITU-T RECMN*E.493 96 W 4862593 ObL73b4 449 W FOREWORD The -T (Telecommunication Standardization Sector) is a permanent organ of the International Telecommunication Union (). The -T is responsible for studying technical, operating and tariff questions and issuing
3、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 on the
4、se topics. The approval of Recommendations by the Members of the ITU-T is covered by the procedure laid down in WTSC Resolution No. 1 (Helsinki, March 1-12, 1993). -T Recommendation E.493 was prepared by -T Study Group 2 (1993-1996) and was approved under the WTSC Resolution No. 1 procedure on the 1
5、9th 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 ITU 1996 All rights reserved. No part of this publication may be reproduced or utilized in any form or
6、 by any means, electronic or mechanical, including photocopying and microfilm, without permission in writing from the ITlJ. 1 2 3 4 5 6 7 8 9 10 ITU-T RECMN*E.493 96 D 4862591 0637365 385 = CONTENTS Scope References Introduction Monitoring network GOS parameters for circuit-switched services in ISDN
7、 Monitoring GOS parameters for mobile services . Monitoring GOS parameters for B-ISDN Monitoring GOS parameters for Signalling System No . 7 networks . History Monitoring GOS parameters for IN-based services . Monitoring GOS in digital international telephone exchanges Page 1 1 2 3 9 10 11 11 11 11
8、Recommendation E.493 (OU96) 1 ITU-T RECMN*E-493 96 4862593 0637366 211 SUMMARY This Recommendation provides the methods for monitoring GOS parameters in fixed and mobile telecommunications networks. GOS monitoring is considered for all traffk related GOS parameters defined in ITU-T Recommendations.
9、The GOS monitoring methodology is based on the principle that the measurements that are available are those that a network operator can obtain on a continuous basis using measurements taken at network elements. All measurements are defined so that a network element can provide the data without requi
10、ring coordination or communication with any other network element. An implication of this principle is that for all GOS parameters that involve multiple networks, each networks contribution to the GOS parameter is monitored and the overall GOS parameter is monitored only if it is possible from measu
11、rements taken within a single network. 11 Recommendation E.493 (02/96) ITU-T RECMN*E.493 9b 9 4862591 0637367 158 9 Recommendation E.493 GRADE OF SERVICE (GOS) MONITORING (Geneva, 1996) 1 Scope This Recommendation covers the monitoring of traffk related Grade of Service (GOS) parameters in fixed and
12、 mobile telecommunications networks. The services considered are circuit-switched services, packet-switched services, Intelligent Network (IN)-based services and Broadband ISDN (B-1SDN)-based services. Many of the GOS parameters defined are end-to-end in nature and to directly measure the parameter,
13、 the measurement equipment would have to be at the customer location (e.g. for post-selection delay and end-to-end blocking), or measurements would have to be taken between two locations e.g. answer signal delay would be measured between two customer locations, Initial Address Message (IAM) delay wo
14、uld be measured between two local exchange switches. This Recommendation does not consider measurements of that type. End-to-end or customer-based measurement usually involves temporary installations of monitoring equipment, and is not done on a continuous basis. Quality of Service (QOS) monitoring
15、is frequently done in this way. The type of GOS monitoring considered in this Recommendation is of a type that a network operator can do on a continuous basis using measurements taken at the network elements. Monitoring with specialized measurement equipment that connects to the network at one or mo
16、re places is not considered. All measurements are defined so that a network element can collect the data without requiring coordination or communication (for the purpose of taking measurements) with any other network element. That is, each network element acts autonomously with regard to taking meas
17、urements. The implication of this restriction on how measurements are taken is that the defined GOS parameters are not measured directly, but rather, each networks contribution to the defined parameter is measured. In order for network operators to assess how well their networks are meeting objectiv
18、es, an allocation for each network needs to be made of the target values for the defined GOS parameters. This Recommendation does not cover how to do this allocation. The allocation will depend on how a specific network interconnects with other networks for different services. It is up to the networ
19、k operators to work out these allocations with their interconnection partners. Recommendation E.721 provides the reference connection models used in determining the target GOS parameter values given there, and those reference connections should be considered in making allocations of target values to
20、 separate networks. 2 References The following CC and ITU-T Recommendations provide the definitions and target values for the GOS parameters. considered in this Recommendation: - - CCIT Recommendation ES43 (1988), Grades of service in digital international telephone exchanges. CC Recommendation E.72
21、1 (1991), Network grade of service parameters and target values for circuit-switched services in the wolving ISDN. CCIT Recommendation E.723 (1992), Grade-of-service parameters for Signalling System No. 7 networks. IT-T Recommendation E.724 (1996), GOS parameters and target GOS objectives for IN ser
22、vices. ITU-T Recommendation E.771 (1993), Network grade of service parameters and target values for circuit-switched land mobile services. ITU-T Recommendation E.774), Network grade of service parameters and target values for maritime and aeronautical mobile services. - - - - ) Presently at the stag
23、e of draft. Recommendation E.493 (0296) 1 ITU-T RECMNtE.473 76 4862593 Ob17368 094 W - -T Recommendation E.7762) , Network grade of service parameters and target values for UPT. The following CCITT and -T Recommendations contain information related to this Recommendation: - CCIT Recommendation ES02
24、(1992), Trafic measurement requirements for digital telecommunication exchanges. CCT Recommendation ES05 (1992), Measurements of the performance of common channel signalling network. CCT Recommendation ES40 (1988), Overall grade of service of the intemutional part of an international connection. - C
25、CITT Recommendation ES41 (1988). Overall grade of service for intemutional connections (subscriber-to-subscriber). -T Recommendation E.743 (1999, Trafic measurements for SS No. 7 dimensioning and planning. -T Recommendation E.770 (1993), Lund mobile and fixed network interconnection traffic grade of
26、 service concept. ITU-T Recommendation E.7732), Maritime and aeronautical mobile COS concept. iTWT Recommendation E.775 (1996), UPT grade of service concept. - - - - - - 3 Introduction Traffic-related GOS parameters are of two general types: delay parameters and probability of mishandling or blocked
27、 call parameters. The probability of steady-state blocking or malfunctioning measurements are usually straightforward. The total number of events (e.g. call attempts) are counted and the number of malfunctions (e.g. blocked calls) are counted, and taking the ratio gives the desired measure. Delay me
28、asures can be more difficult. One problem is that many delay parameters measure delays between different points in a network. It is desirable to take direct measurements whenever possible, but sometimes this cannot be done. The SS No. 7 protocol does not provide for time stamping of messages and so
29、a direct measurement of delay using time stamps on the actual message is not possible. One of the basic principles of GOS monitoring that is taken in this Recommendation is that no coordination should be required between network elements to take the measurements. Each network element should work aut
30、onomously in taking measurements. This means that each network element needs to take delay measures within itself. The combination of individual network element performances to estimate defined end- to-end GOS parameters can be approximated using reference connections such as those given in clause 3
31、E.721. Another difficulty with delay measures is that frequently 95% target values are specified. Since for some GOS parameters the methods used here do not measure the GOS parameter directly, but rather add and subtract different measures to arrive at the GOS parameter, the 95% calculations can be
32、troublesome. To do these calculations accurately, correlations of Probability distributions are required, and this can become complicated and require significant data storage. An alternative suggested here is to use the following approximation. Approximation for computing 95% values The approximatio
33、n is to assume that for any random variable considered, say z, its mean, Z and its 95% value, 295, are related by: where oz is the standard deviation of z. The important part of the assumption is that a is the same for all random variables considered. One class of random variables satisfying this co
34、ndition are those having a normal distribution. 2, Presently at the stage of draft. 2 Recommendation E.493 (02/96) ITU-T RECMN*E*493 9b 4862573 Ob37367 T20 Consider independent random variables xi for which the mean values, Xi and 95% values, xp5, are known. Note that these are easily computed from
35、data and do not require distribution information to be stored. If the Xi are assumed to satisfy the above condition, then if N i=l z= cxi il sensuit que: N N 112 z95 = Fi + c. (xps - x.” i=l i=t The ab ve approximati n is u ed in this Recommendation to derive expressions for 95% values of sums of in
36、d pendent random variables. The above expression is derived using the fact that the variance of a sum of independent random variables is the sum of their variances. 4 Monitoring network GOS parameters for circuit-switched services in ISDN The following network GOS parameters for circuit-switched ser
37、vices in ISDN are defined in clause 2E.721: pre-selection delay (overlap sending); post-selection delay (overlap sending); post-selection delay (en bloc sending); 0 answer signal delay; call release delay; and 0 probability of end-to-end blocking. The target values for these GOS parameters are given
38、 in clause 3E.721. An individual circuit-switched network is defined as a collection of interconnected network elements (e.g. switches, service control points, signal transfer points, etc.) that: 0 has the capability to establish a connection between a subscriber or network and another subscriber or
39、 network; and 0 is operated by a single network operator. Figure 1 illustrates the idea of individual networks and local, toll and international end-to-end connections. Network 1 is a local area network that provides local connections (e.g. between T1 andT), toll connections (e.g. for a toll call fr
40、om Ti to T3, network 1 provides a connection between T1 and network 2) and intemational connections (e.g. for an international call from T1 to T4, network 1 provides a connection between T1 and network 2). Network 2 provides toll connections (e.g. for the call from T1 to T3, it provides a connection
41、 between network 1 and network 3) and international connections (e.g. for the call between T1 and T4, it provides the connection between network 1 and network 4). There are many possibilities for different network configurations and interconnections. The principle used below in monitoring the above
42、GOS parameters is that monitoring should be done by each individual switched network to determine its contribution to the end-to-end GOS parameters. Target values for each individual networks contribution to a GOS parameter must be determined using reference connection models that reflect the role t
43、hat network plays in different connection types. Recommendation E.493 (02196) 3 ITU-T RECMN*E=493 96 4862573 Ob37370 742 Network 2 FIGURE 1E.493 illustration of networks and end-bend connectim 4.1 Monitoring delay GOS parameters A signal flow diagram for a circuit-switched connection in ISDN is show
44、n in Figure 2. The four delay parameters identified above are denoted by tl to r4 as follows: tl Re-selection delay (overlap sending). t2 Post-selection delay (overlap or en bloc sending). t3 Answer signal delay. t4 Call release delay. These four delays, rl to t4 , are identified in Figure 2. It is
45、clear that tl, t2 and r4 are directly measurable only from the calling terminal (A). The answer signal delay is not directly measurable from any single point in the connection, Since the objective of GOS monitoring is for each network operator to be able to take measurements from their network and a
46、ssess how their network is performing relative to the GOS objectives, a methodology other than direct measurement of the defined GOS parameters is required for monitoring the delay GOS parameters. The following subclauses define the methodology that should be used and the measurements required for e
47、ach network operator to monitor their contribution to each of the delay GOS parameters tl through t4. 4.1.1 Monitoring pre-selection delay tl The components of the pre-selection delay rl are the emission time and propagation time for the Set-up message from the calling terminal to the local exchange
48、 serving the calling terminal, the intra-office delay ft (see Figure 2) at the local exchange, the emission time and propagation time for the set-up ACK message from the local exchange to the calling terminal. The component that is measurable by a network switch is ti. Therefore, the methodology rec
49、ommended is that, for each local exchange area, the mean and 95% values of ff are determined for the local exchange area in the manner described below. Each of those values are then added to 2 x (maximum propagation time from terminal to local exchange) + Set-up message emission time) + Set-up ACK message emission time. This gives an estimate of the mean and 95% for tl that is conservative since a worst case has been used for propagation delays. The variability in propagation delay within a local exchange area is small enough that this a
