1、CEPT T/GSI*03-08/ L 85 232b43i.1 OOOSbO 4 I T/SI 03-0811 E Page 1 Recommendation T/GSI OS-08/1 wice 1985, revised by correspondence, June 1987) NETWORK PERFORMANCE IN THE ISDN Recommendation proposed by the Special Group Integration (GSI) Revised text of the Recommendation adopted by the “Telecommun
2、ications” Commission: “The European Conference of Posts and Telecommunications Administrations, considering - that the networks technical performance needs to be defined with respect to - end to end connections - connection elements - basic connection components as specified in Recommendation T/GSI
3、03-05 with the title “Reference configuration for ISDN connection types”, recommends the application of the following definitions on the general aspects of network performance in the ISDN and its structuring.” Note. The detailed definition of network performance parameters is the subject of separate
4、 Recommendations. - I -7 Edition of October 31, 1987 ICEPI T/GSI*03-08/ LE 5 2326434 0005601 b TIGSI 03-0811 E Page 2 1. GENERAL DESCRIPTION OF NETWORK PERFORMANCE 1.1. Network performance in the ISDN has to be considered as applicable to different connection types. Each connection type or group of
5、connection types may be described by an individual set of attributes. One of these attributes deals with performance parameters. Recommendation T/GSI 03-04, which describes and specifies the ISDN connection types, allows for so many variations in the values of its different attributes which leads to
6、 a very large number of potential connection types, that it is necessary to consider only certain dominant attributes in order to produce a simpler list of reference congurations. For an initial analysis only the first two of the four dominant attributes in Recommendation T/GSI 03-04 need to be cons
7、idered. Therefore the “information transfer mode” and “information transfer rate” attributes will lead to three general groups of ISDN connection types, viz. : - 64 kbit/s circuit, - packet, - broadband circuit ( 64 kbit/s). The other two dominant attributes (information transfer capability and esta
8、blishment of connection) do not require separate reference configurations but will manifest themselves by different performance values. 1.2. Network performance may be considered in five basic groups as follows: i) connection processing delays; ii) connection failures due to congestion (blocking); i
9、ii) connection failures due to malfunction; iv) loss of network capability; v) transmission performance. In telecommunications networks it is necessary for economic reasons to limit the resources provided for carrying the offered trafic. This limitation may affect the quality of service to the user
10、of switched ISDN connections in two different ways: by call processing delays and/or by blocking. In order to allocate network performance parameters it is necessary to have a reference model over which the parameters can apply and which also allows a mechanism for allowing allocations of sub-divisi
11、ons of the value of a parameter to be made between different sections of an ISDN connection. This reference model is called “Reference Conjgurntions for ISDN Connection Types” (Recommendation T/GSI 03-05). 1.3. 2. QUALITATIVE DESCRIPTION OF NP PARAMETERS Paragraph 1.2. above already contains a struc
12、tural approach as a basis for further study on NP parameters. This approach has been assumed for the proposal of a list of parameters which is contained in Table 1 (T/GSI 03-OS/l) below. i) within each group of parameters there are a small number of NP parameter primitives which, generally speaking,
13、 are independent of the connection type involved, but which take into account the different phases of a call as defined in the OS1 Network Layer Service definition; the actual Network Performance Parameters are associated with these primitives. ii) These Network Performance Parameters can be used to
14、 describe the end-to-end performance of a connec- tion, individual parts of an ISDN Connection type (connection element), or in some cases Basic Connection components (see Recommandation T/GSI 03-03 section 4.2.3.). An individual performance parameter may be applicable to each of the three situation
15、s above or may be applicable to only one or two of them. Note. It is necessary to have end-to-end performance parameters and their values, specified since this is what is used to map Network Performance to Quality of Service parameters and their values. (The exact Relationship between these two type
16、s of parameters needs to be further elaborated taking into account design objectives, operational performance objectives, “network affects” and in the case of teleservices the TE and customer network performance.) Performance parameters for connection elements need to be specified in cases where the
17、re is a need to specify equipment performance parameters on a type of connection basis (e.g. local, national trunk, or international connection); or where the ISDN is in fact only providing some of the connection elements in interworking situations into dedicated networks; or where there is a need t
18、o allocate portions of the overall network connection between different countries or different network operators. - Edition of October 31, 1987 CEPT T/GSI*03-08/ ZE 85 23264Z1i 0005602 .8 W 1 NF Parameter Groups (Note 3) T/GCI 03-0811 E Page 3 NP Parameter Specic Primitives NP Parameters (Note 4) (y
19、et to be defined) Performance parameters for basic connection components (e.g. exchange connection-see Recommenda- tions T/CS 62-04 and T/CS 68-04) have already been developed for a number of components (e.g, digital muldex equipment, digital line sections, exchange connections). The first purpose i
20、n developing these parameters and agreed values is to assist equipment manufacturers and purchasers (e.g. Administrations and network operators) in having standardized equipment available at lower costs. The second reason is that these equipments will naturally be assembled to form a national ISDN c
21、onnection type or connection elements. It is therefore possible by this bottom-up approach to arrive at performance objectives of the connection type or connection element for a particular known national implementation plan, which should conform to agreed architectural descriptions (Reference config
22、urations) for that connection type. This can be done by a summation process on the performance of the connection components together with some elaborations due to design/operational reasons. Due to the normal situation of international standards having to be a compromise between the economic benefit
23、 of having standardized operation and conditions leading to greater traffic and lower costs on the one hand, and the need to satisfy national conditions of varying geography, particular technologies, regulatory conditions, economic circumstances on the other, it is possible that the bottom-up and to
24、p down approaches may diverge in the actual values of parameters. The whole process, for each approach, has to be iteratively studied to achieve the desired goals of the different aspects. Connection Processing Delays Study Areas Info. transfer Delay Trafficability Performance Availability Performan
25、ce Transmission Performance I (see Conn. Delays Recommendation 03-0811) Disc. Delays (see I Recommendation 03-OS/l) Com. failure Connection Failures due to congestion Conn. failure Disc. failure Connection Failures due to malfunction (Note 1) Loss of Network Capability (Note 2) Transmission Performa
26、nce Occurrence, Duration Error Performance Jitter and wander Controlled slip Short interruption on connections Timing and synchronisation Propagation delay of the transmission Note I. This is a failure affecting a single connection or phase of a connection. Note 2. This parameter impacts on the avai
27、lability of individual network elements, which could have an impact on call attempts and calls for a specific number of subscribers. Note 3. These parameter groups may be applicable only to network components, at which they are measurable and distinguishable. Note 4. The list of NE-parameter primiti
28、ves indicate the overall list of NP-parameters which need to be developed. It is possible that some of the parameters will interact with each other. This is particularly by so in the case of the relationship between the connection related parameters and the Loss of Network Capability parameters. Tab
29、le 1 (T/GSI 03-08/1). Qualitative Description of Network Performance Parameters. f .- 7 Edition of October 31, 1987 I i I 1 i t ICEPT T/GSI*03-0B/ LE 85 2326434 0005b03 T m T/GSI 03-0811 E Page 4 3. QUANTITATIVE DESCRIPTION OF NP PARAMETERS For each of the NP parameters yet to be defined a number of
30、 numerical values expressing design objectives need to be specified. The number of values which may be assigned to each parameter will depend on the following facts: i) the assumed reference configuration (end-to-end connection or connection element, connection of moderate length or longest connection); ii) mean value to be specified and 95 per cent value; iii) values specified for normal busy hour load or high load conditions. Certain parameters and their associated values may be influenced by only a part of the assumptions listed above. - Edition of October 31, 1987
