ITU-T E 437-1999 Comparative Metrics for Network Performance Management - Series E Overall Network Operation Telephone Service Service Operation and Human Factors Quality of Servic.pdf

1、INTERNATIONAL. TELECOMMUNICATION UNION ITU-T TELECOMMUNICATION STANDARDIZATION SECTOR OF ITU E.437 (05/99) SERIES E: OVERALL NETWORK OPERATION, TELEPHONE SERVICE, SERVICE OPERATION AND HUMAN FACTORS Quality of service, network management and traffic engineering - Network management - Checking the qu

2、ality of the international telephone service Comparative metrics for network performance management ITU-T Recommendation E.437 (Previously CCITT Recommendation) STD*ITU-T RECMN E.437-ENGL 1999 Measurement and recording of traffic Forecasting of traffic Determination of the number of circuits in manu

3、al operation Determination of the number of circuits in automatic and semi-automatic operation Grade of service Definitions ISDN traffic engineering Mobile network trafic engineering II 4Bb25L ObbLLb 2Lb ITU-T E-SERIES RECOMMENDATIONS OVERALL NETWORK OPERATION, TELEPHONE SERVICE, SERVICE OPERATION A

4、ND HUMAN FACTORS OPERATION, NUMBERING, ROUTING AND MOBILE SERVICES INTERNATIONAL OPERATION Definitions General provisions concerning Administrations General provisions concerning users Operation of international telephone services Numbering plan of the international telephone service International r

5、outing plan Tones in national signalling systems Maritime mobile service and public land mobile service OPERATIONAL PROVISIONS RELATING TO CHARGING AND ACCOUNTING IN THE INTERNATIONAL TELEPHONE SERVICE I Charging in the international telephone service Measuring and recording call durations for accou

6、nting purposes (1 E. 1 OO-E. 103 E.104-E. 1 19 E. 120-E. I 39 E. 140-E. 159 E. 160-E. 169 E. 170-E. 179 E. 1 80-E. 199 E.200-E.229 E.230-E.249 E.260-E.269 UTILIZATION OF THE INTERNATIONAL TELEPHONE NETWORK FOR NON- TELEPHONY APPLICATIONS General E.300-E.319 Phototelegraphy E.320-E.329 E. 330-E. 399

7、ISDN PROVISIONS CONCERNING USERS QUALITY OF SERVICE, NETWORK MANAGEMENT AND TRAFFIC ENGINEERING NETWORK MANAGEMENT International service statistics E.400-E.409 International network management E.41 0-E.419 TRAFFIC ENGINEERING / E.490-E.505 E.506-E.509 E. 5 i i ASR is useful as a comparative metric.

8、When examining the performance of multiple routes to common destinations, any difference in ASR should be attributed directly to the networks involved. Care must be taken to ensure that proper comparisons are conducted as described in clause 5 (Remarks). t Answer supervision TO2076w-98 3.2 The speed

9、 with which a network responds to a user who is requesting that a connection be established is an important quality factor that is readily discernible by customers. Recommendation E.43 1 (Service quality assessment for connection set-up and release delays) defines three relevant time intervals: star

10、t dial signal delay, post dialling delay and call clearing delay. Of these, post dialling delay (PDD) provides a view of the time it takes networks to establish connections after the customer has completed dialling the destination address. PDD is most often measured on an end-to-end basis, requiring

11、 test equipment be placed at call originating and terminating points. PDD may also be measured by the observation of appropriate signalling messages within networks. A metric that allows call Set-up time on multiple routes to the same destination to be compared would provide a cost efficient method

12、of assessing the quality on those routes. Post gateway answer delay (PGAD) is such a metric. PGAD is defined for calls that are answered as follows: . Post gateway answer delay (PGAD) PGAD is the time interval between the seizure of the international circuit and the receipt of answer supervision. Wi

13、thin this interval, the time between the seizure and the first network response is a function of the performance of the network, while the time between the network response and the answer is user driven. When looked at for a single route, PGAD would not be terribly useful. However, when comparing tw

14、o routes with large and carefully selected data samples, there should be no significant difference in customer behaviour between the two routes and any significant difference in PGAD could be attributed to the networks involved. PGAD, and its relationship to PDD is illustrated in Figure 1. Dial c 4

15、PGAD b Ringing 4 3.3 Another parameter that provides useful data on a comparative basis is average length of conversation (ALOC) for completed calls. When comparing routes to the same destination, where each route carries a portion of a common traffic stream, it should be expected that average conve

16、rsation times on each route would be comparable. A statistically significant difference in ALOC between two routes could be considered as an indication of some irregularity warranting further investigation. Average length of conversation (ALOC) Recommendation E.437 (05/99) 3 ALOC is measured for com

17、pleted calls only. Preferably, ALOC should be measured from the time of answer-supervision to the time of call release. If measurement systems do not support this, ALOC may be measured from the time of trunk seizure. In this case, call set-up time is included in the measure and variations in the cal

18、l Set-up time could have an influence on ALOC. However, because call Set-up time is typically small compared to conversation time, any error introduced should be small. Regardless of where the ALOC measurement interval begins, all routes being measured should be measured the same way. Several factor

19、s could impact ALOC. A route that is subjected to an increased level in network- caused call cut-Offs would have a lower ALOC than the reference route. The same would hold true if there was a greater frequency of facsimile transaction failures. The use of double-hop satellite or tandem compression e

20、quipment on a route may reduce voice transmission quality that could lead to a reduction in ALOC. Signalling-related problems and numbering plan changes could lead to short holding time calls that would impact ALOC. Additional factors other than those indicated here could result in a difference in A

21、LOC between two routes. Care must be taken to ensure that proper comparisons are conducted as described in clause 5 (Remarks). 4 Sta tis tical comparisons A typical application of PGAD and ALOC data would involve comparing the mean value for the metric obtained for one route against the mean value o

22、btained for the baseline route. When doing such comparisons, it is important to employ some statistical mechanisms to determine if any observed difference is statistically significant. Factors affecting the significance between two sample means include the dispersion or distribution of the samples u

23、sed to compute the mean along with the number of samples obtained. The formula given below is an example of how statistical methods could be applied to PGAD data. In this example, we want to test if the PGAD on one route exceeds that of a baseline route by some factor. A one tailed test is performed

24、 to test this comparison to a 95% confidence level. PA = Mean PGAD for route under study (for one measurement interval) PB = Mean PGAD for baseline route (for one measurement interval) oA = Standard deviation of PGAD for route under study oB = Standard deviation of PGAD for baseline route NA = Numbe

25、r of calls on route under study NB = Number of calls on baseline route If D exceeds some value (X), then it can be said that with 95% confidence, the PGAD on the route under study exceeds that of the baseline route by X seconds. NOTE - If standard deviation is not available, then a value of one half

26、 the mean value may be used as a conservative estimate. - 4 Recommendation E.437 (05/99) STD*ITU-T RECMN E-437-ENGL 1999 48b259L ObbL177 392 I 5 Remarks Unless otherwise noted, the following remarks apply to all metrics used in a comparative manner. 1) 2) When using any comparative measure, care mus

27、t be taken so that the routes being compared carry identical mixtures of services and serve the same completing fields. Comparative measures are useful when comparing multiple routes to the same destinations. Routing mechanisms such as code or carrier specific routing must be taken into account when

28、 comparing the performance of multiple routes. Data for comparative measures may be obtainable fi-om switch call detail records. This method of data collection as well as the collection of data fi-om other suitable network monitoring sources is often very efficient, providing large samples at relati

29、vely low incremental cost. Sample sizes must be suficiently large to ensure statistical integrity of the data. Statistical means to test the significance of any difference in any comparative measure between different routes must be used. The availability of the appropriate measures of central tenden

30、cy and dispersion along with the corresponding standard error estimates is desirable. When a significant difference in ALOC is identified, detailed analysis should be undertaken to determine the root cause so that corrective action can be taken. 3) 4) 5) 6) .* Recommendation E.437 (05/99) 5 STD-ITU-

31、T RECMN E.437-ENGL 1999 48b2591 Obb1178 229 ieries A leries B leries C leries D ieries E ;cries F ;eries G Series H Series I Series J Series K Series L Series M Series N Series O Series P Series Q Series R Series S Series T Series U Series V Series X Series Y Series Z ITU-T RECOMMENDATIONS SERIES Or

32、ganization of the work of the ITU-T Means of expression: definitions, symbols, classification General telecommunication statistics General tariff principles Overall network operation, telephone service, service operation and human factors Non-telephone telecommunication services Transmission systems

33、 and media, digital systems and networks Audiovisual and multimedia systems Integrated services digital network Transmission of television, sound programme and other multimedia signals Protection against interference Construction, installation and protection of cables and other elements of outside p

34、lant TMN and network maintenance: international transmission systems, telephone circuits telegraphy, facsimile and leased circuits Maintenance: international sound programme and television transmission circuits Specifications of measuring equipment Telephone transmission quality, telephone installat

35、ions, local line networks Switching and signalling Telegraph transmission Telegraph services terminal equipment Terminals for telematic services Telegraph switching Data communication over the telephone network Data networks and open system communications Global information infrastructure and Internet protocol aspects Languages and general software aspects for telecommunication systems - Printed in Switzerland Geneva, 1999