a) Error performance Rec. G.821 Rec. G.821 Rec. G.821
.T&
lw(48p) | lw(60p) | lw(60p) | lw(60p) .
b) Slip performance Rec. G.822 Rec. G.822 Rec. G.822
.T&
lw(48p) | lw(60p) | lw(60p) | lw(60p) .
12 Network interworking FS FS FS
.T&
lw(48p) | lw(60p) | lw(60p) | lw(60p) .
13 Operations and management FS FS FS
.TE
.LP
FS\ further study
.LP
\ua\d\u)\d
Where there are two or more S/T interfaces, different values of
access attributes (attributes 8, 9 and 10) may occur at each interface. Values need to be specified for each channel of the interface structure. The role of the access attributes in determining connection types is for further study.
Interfaces to network specialized resources and to other networks are for
furter study.
.LP
\ub\d\u)\d
Examples of the additional performance attributes which may be
defined are:
.LP
\(em
call and packet processing delays;
.LP
\(em
probability of call faillure due to congestion;
.LP
\(em
probability of call failure due to network malfunction or
packet mishandling;
.LP
\(em
information transfer delay;
.LP
\(em
error performance [including attributes 11 a) and 11 b)].
.LP
\uc\d\u)\d
Use of Rec. X.25 and X.75 in ISDN can be found in
Rec.\ X.31.
.LP
\ud\d\u)\d
Packet connection establishment/release may be a two stage
process: stage 1 the selection of a B\(hychannel, stage\ 2 the setting
up of a packet connection. For further details, see
Rec.\ X.31.
.nr PS 9
.RT
.ad r
\fBTableau 1/I.340 [2T1.340], p.6\fR
.sp 1P
.RT
.ad b
.RT
.LP
.bp
.sp 2P
.LP
3.2.1
\fIInformation transfer mode\fR
.sp 1P
.RT
.sp 1P
.LP
\fIAttribute values for connection elements\fR
.sp 9p
.RT
.LP
Circuit or packet.
.sp 1P
.LP
\fIAttribute values for overall connection type\fR
.sp 9p
.RT
.LP
Circuit or packet.
.sp 1P
.LP
\fIAssociation law\fR
.sp 9p
.RT
.LP
Due to the nature of current packet systems, the use of packet
mode in any connection element would make the overall connection type
a packet type.
.sp 2P
.LP
3.2.2
\fIInformation transfer rate (kbit/s)\fR
.sp 1P
.RT
.sp 1P
.LP
\fIAttribute values for connection elements\fR
.sp 9p
.RT
.LP
16 or 32 or 64 or 2\ \(mu\ 64 or 384 or 1536 or 1920
.LP
(Values 16 and 32 are not allowed in the access connection
element).
.sp 1P
.LP
\fIAttribute values for overall connection type\fR
.sp 9p
.RT
.LP
(16 or 32) or 64 or 2\ \(mu\ 64 or 384 or 1536 or 1920.
.sp 1P
.LP
\fIAssociation law\fR
.sp 9p
.RT
.LP
The value for the overall connection type will be equal to the
equipment, speech interpolation, \(*m/A\ law conversion) and/or echo
suppression functions and/or multiple satellite hops or
null.
.LP
The exact means of specification of the attribute is for further
study. One method would be an appropriate reference to a
Recommendation detailing operational requirements in the
ISDN.
.sp 1P
.LP
\fIAttribute values for overall connection types\fR
.sp 9p
.RT
.LP
Unrestricted digital information or 3.1\ kHz audio or
speech.
.sp 1P
.LP
\fIAssociation law\fR
.sp 9p
.RT
.LP
For an overall connection type to have the value \fIunrestricted\fR
\fIdigital\fR , no connection element may contain speech processing
functions or echo suppression functions. Connection elements
containing speech processing devices having the flexibility to change
operation between speech and 64\ kbit/s unrestricted would on the other
hand be allowed to be a part of a number of different connection
types.
.LP
For an overall connection type to have the value \fI3.1 kHz audio\fR ,
it may contain echo suppression functions (or it has to disable them
prior to data transfer); it must however contain \(*m/law\ conversion
equipment when appropriate.
.LP
For an overall connection type to have the value \fIspeech\fR , it must
contain \(*m/law\ conversion equipment and echo suppression functions when
appropriate.
.LP
These matters are dealt with in more detail in
Recommendation\ I.335.
.bp
.sp 2P
.LP
3.2.4
\fIEstablishment of connection\fR
.sp 1P
.RT
.sp 1P
.LP
\fIAttribute values for connection elements\fR
.sp 9p
.RT
.LP
Switched or semi\(hypermanent or permanent.
.sp 1P
.LP
\fIAttribute values for overall connection type\fR
.sp 9p
.RT
.LP
Switched or semi\(hypermanent or permanent.
.sp 1P
.LP
\fIAssociation law\fR
.sp 9p
.RT
.LP
If all connection elements are permanent, then the overall
connection type is permanent.
.LP
If any of the connection elements are switched, then the overall
connection type is switched. If one or more of the connection elements
are semi\(hypermanent and none of the connection elements are switched,
then the overall connection type is semi\(hypermanent.
.sp 2P
.LP
3.2.5
\fISymmetry\fR
.sp 1P
.RT
.sp 1P
.LP
\fIAttribute values for connection elements\fR
.sp 9p
.RT
.LP
Unidirectional or bidirectional symmetric or bidirectional
asymmetric.
.sp 1P
.LP
\fIAttribute values for the overall connection type\fR
.sp 9p
.RT
.LP
Unidirectional or bidirectional symmetric or bidirectional
asymmetric.
.sp 1P
.LP
\fIAssociation law\fR
.sp 9p
.RT
.LP
The overall symmetry can only be generated from the connection
elements by analysis of the connection element values in the context
of the architecture of the connection.
.sp 2P
.LP
3.2.6
\fIConnection of configuration\fR
.sp 1P
.RT
.sp 1P
.LP
3.2.6.1
\fITopology\fR
.sp 9p
.RT
.sp 1P
.LP
\fIAttribute values for connection elements\fR
.sp 9p
.RT
.LP
Point\(hyto\(hypoint (simple, tandem or 2\ \(mu\ 64 parallel), or
multipoint.
.LP
(The access connection element may not be
multipoint.)
.sp 1P
.LP
\fIAttribute values for the overall connection type\fR
.sp 9p
.RT
.LP
Local or national or international. (Each simple or 2\ \(mu\ 64
parallel.)
.sp 1P
.LP
\fIAssociation law\fR
.sp 9p
.RT
.LP
No association is possible.
.sp 2P
.LP
3.2.6.2
\fIUniformity\fR
.sp 1P
.RT
.sp 1P
.LP
\fIAttribute values for connection elements\fR
.sp 9p
.RT
.LP
Uniform or non\(hyuniform.
.sp 1P
.LP
\fIAttribute values for the overall connection type\fR
.sp 9p
.RT
.LP
Not applicable.
.sp 1P
.LP
\fIAssociation law\fR
.sp 9p
.RT
.LP
Not applicable.
.bp
.sp 2P
.LP
3.2.6.3
\fIDynamics\fR
.sp 1P
.RT
.sp 1P
.LP
\fIAttribute values for connection elements\fR
.sp 9p
.RT
.LP
Not applicable.
.sp 1P
.LP
\fIAttribute values for the overall connection type\fR
.sp 9p
.RT
.LP
Concurrent or sequential or add/remove, or symmetry and/or
topology change.
.sp 1P
.LP
\fIAssociation law\fR
.sp 9p
.RT
.LP
Not applicable.
.sp 2P
.LP
3.2.7
\fIStructure\fR
.sp 1P
.RT
.sp 1P
.LP
\fIAttribute values for connection elements\fR
.sp 9p
.RT
.LP
Layer\ 1:
8\ kHz integrity or 8\ kHz integrity with RDTD
(Restricted Differential Time Delay)
.FS
The term
\fBRTTD\fR in the connection type context is defined as
follows:
This value applies when:
i)
at each point in a connection or connection element, the
time slots are explicitly or implicitly demarcated for each
information channel or an aggregate of information channels,
and
.LP
ii)
the information parts submitted to the time slots at the
the transmitting end are delivered to the receiving end with a
differential time delay or not more than 50\ ms.
.FE
\u,\d |
.FS
50 ms
is a provisional value that needs to be confirmed. This value has to
take into account the maximum differential time delay of an
appropriate HRX or part thereof as defined in the G\(hySeries
Recommendations.
.FE
or unstructured
.LP
Layer\ 2:
Service data integrity or
unstructured
.LP
Layer\ 3:
Service data integrity or
unstructured
.sp 1P
.LP
\fIAttribute values for the overall connection type\fR
.sp 9p
.RT
.LP
As per values for connection elements.
.sp 1P
.LP
\fIAssociation law\fR
.sp 9p
.RT
.LP
For further study.
.sp 2P
.LP
3.2.8
\fIChannels\fR
.sp 1P
.RT
.sp 1P
.LP
3.2.8.1
\fIInformation channel (rate)\fR
.sp 9p
.RT
.PP
\fIAttribute values for connection elements\fR
.RT
.LP
Access\ connection\ element:
D(16) or D(64) or B(64) or
H\d0\u(384) or H\d1\\d1\u(1536) or
H\d1\\d2\u(1920)
.LP
Transit\ connection\ element:
64\ kbit/s or equivalent in a
higher order multiplex or packet system
or analogue transmission.
.sp 1P
.LP
\fIAttribute values for the overall connection type\fR
.sp 9p
.RT
.LP
Not applicable
.sp 1P
.LP
3.2.8.2
\fISignalling channel (rate)\fR
.sp 9p
.RT
.PP
\fIAttribute values for connection elements\fR
.RT
.LP
Access\ connection\ element:
D(16) or D(64) or B(64)\ +\ D(16) or
B(64)\ +\ D(64)
.LP
Transit\ connection\ element:
common channel signalling system
or packet
.sp 1P
.LP
\fIAttribute values for the overall connection type\fR
.sp 9p
.RT
.LP
Not applicable
.bp
.sp 1P
.LP
3.2.9
\fIConnection control protocol\fR
.sp 9p
.RT
.PP
\fIAttribute values for connection elements\fR
.PP
Access connection element:
.RT
.LP
Layer\ 1:
I.430 or I.431
.LP
Layer\ 2:
I.441 or I.441 + X.25
link level
.LP
Layer\ 3:
I.451 or I.451 + X.25
packet level
.LP
Transit connection element:
.LP
Layer\ 1:
Q.702 or X.75 physical level
.LP
Layer\ 2:
Q.703 or X.75 link level or Q.703 + X.25
link level
.LP
Layer\ 3:
Q.704 + SCCP or Q.704 + ISUP or X.75 packet level or
Q.704 + SCCP + X.25 packet level or
Q.704 + ISUP + X.25 packet level
.sp 1P
.LP
\fIAttribute values for the overall connection type\fR
.sp 9p
.RT
.LP
Not applicable.
.sp 1P
.LP
3.2.10
\fIInformation transfer coding/protocol\fR
.sp 9p
.RT
.PP
\fIAttribute values for connection elements\fR
.PP
Access connection element:
.RT
.LP
Layer\ 1:
I.430 or I.431 or I.430 + G.711 or
I.431 + G.711
.LP
Layer\ 2:
I.441 or X.25 link level or null
.LP
Layer\ 3:
I.451 or X.25 link packet level or null
.LP
Transit connection element:
.LP
Layer\ 1:
G.711 or G.702 or X.75 physical level
.LP
Layer\ 2:
Q.703 or X.25 link level or X.75
link level or null
.LP
Layer\ 3:
X.25 packet level or X.75 packet level or
Q.704 + ISUP or null
.sp 1P
.LP
\fIAttribute values for the overall connection type\fR
.sp 9p
.RT
.LP
Not applicable.
.sp 2P
.LP
3.2.11
\fINetwork performance\fR
.sp 1P
.RT
.sp 1P
.LP
3.2.11.1
\fIError performance\fR
.sp 9p
.RT
.sp 1P
.LP
\fIAttribute values for connection elements\fR
.sp 9p
.RT
.LP
G.821
.sp 1P
.LP
\fIAttribute values for the overall connection type\fR
.sp 9p
.RT
.LP
G.821
.sp 1P
.LP
\fIAssociation law\fR
.sp 9p
.RT
.LP
G.821
.sp 2P
.LP
3.2.11.2
\fISlip performance\fR
.sp 1P
.RT
.sp 1P
.LP
\fIAttribute values for connection elements\fR
.sp 9p
.RT
.LP
G.822
.sp 1P
.LP
\fIAttribute values for overall connection type\fR
.sp 9p
.RT
.LP
G.822
.sp 1P
.LP
\fIAssociation law\fR
.sp 9p
.RT
.LP
G.822
.bp
.sp 1P
.LP
3.2.12
\fIFurther attributes and attribute values\fR
.sp 9p
.RT
.PP
Section 3.2 has outlined the relationships between those attributes values
presently existing; the possibility for new values being added
remains.
.RT
.sp 1P
.LP
3.3
\fILimited set of ISDN connection types\fR
.sp 9p
.RT
.PP
From the given list of attributes and their possible values, a
large number of connection types can be identified. However, some of these
attributes are of a general or dominant nature and an initial set of ISDN
connection types can be based on these dominant attributes.
.PP
Table 2/I.340 enumerates a limited set of connection types
based on the following dominant attributes: information transfer mode,
information transfer rate, information transfer susceptance, establishment
of connection and symmetry. These connection types are intended to be sufficient
to support the basic telecommunication services identified in the
I.200\(hySeries of Recommendations. Additional connection types are for
further study.
.RT
.sp 2P
.LP
\fB4\fR \fBConnection elements\fR
.sp 1P
.RT
.PP
The ISDN network architecture Recommendation\ I.324 explains how an ISDN
connection type is made up of connection elements (CEs). This concept is
illustrated in Figure\ 5/I.340 and this is valid for all connection types
between S/T\ reference points. A particular ISDN connection may be local
(i.e.\ only access connection elements are involved), national transit
(i.e.\ involving access and national transit CEs), or international
(i.e.\ involving all three kinds of CEs).
.PP
Current Recommendations allow collocation and non\(hycollocation of each
of the types of CRFs indicated in Figure\ 5/I.340. This is a national
matter.
.RT
.sp 1P
.LP
4.1
\fIAccess connection element\fR
.sp 9p
.RT
.PP
The access connection element is the portion of the connection from the
S/T reference point to the local connection related function (CRF). In
the case of permanent connection types an equivalent point to the local
CRF needs to be defined.
.RT
.sp 1P
.LP
4.2
\fINational transit connection element\fR
.sp 9p
.RT
.PP
The national transit connection element is the portion of the
connection between the local CRF and the international CRF. In the case of a
national connection this would default to a \*Qtransit connection element\*U,
i.e.\ between two local CRFs, but could involve network elements from more
than one network operator.
.RT
.sp 1P
.LP
4.3
\fIInternational connection element\fR
.sp 9p
.RT
.PP
The international connection element is the portion of the
connection between the originating and destination international
CRFs.
.RT
.sp 1P
.LP
4.4
\fIUse of connection elements\fR
.sp 9p
.RT
.PP
By using connection elements and attributes which have a layered
nature, the construction of a connection type is more easily described.
The use of different values for the same attribute in different connection
elements
allows for a greater degree of description and flexibility.
.PP
The connection element analysis may assist in the description of a
complex and asymmetric ISDN connection. This is illustrated in Figure\
6/I.340, in which the configuration attributes of topology, uniformity
and dynamics
for a connection type are described using the concept of connection
elements.
.PP
Different connection elements which constitute an ISDN connection may have
different sets of attributes. In this case the attributes across the
connection are not homogeneous, and the available attributes of the connection
are limited by the most restrictive set of attributes of all the connection
elements of the connection.
.bp
.RT
.ce
\fBH.T. [1T2.340]\fR
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(342p) .
TABLE\ 2/I.340
.T&
cw(342p) .
{
\fBSet of ISDN connection types\fR
}
.TE
.TS
center box;
cw(36p) | cw(171p) sw(135p) , ^ | c | c.
ISDN connection type identity Attributes
{
Dominant attributes which define ISDN connection types
} Additional attributes
_
.TE
.TS
center box ;
cw(12p) | cw(24p) | cw(24p) | cw(18p) | cw(36p) | cw(27p) | cw(36p) | cw(30p) | cw(18p) | cw(24p) | cw(15p) | cw(21p) | cw(21p) | cw(21p) | cw(15p) , ^ | ^ | l | l | l | l | l | l | l | l | l | l | l | l | l.
Dominant attributes which define ISDN connection types
} Additional attributes
_
.TE
.TS
center box;
cw(12p) | cw(24p) | cw(24p) | cw(18p) | cw(36p) | cw(27p) | cw(36p) | cw(30p) | cw(18p) | cw(24p) | cw(15p) | cw(21p) | cw(21p) | cw(21p) | cw(15p) , ^ | ^ | l | l | l | l | l | l | l | l | l | l | l | l | l.
C 3 Circuit 384 Unrestricted Permanent Unidirectional | ub\d\u)\d Pt\(hypt multipoint 8 kHz H 0(384)
_
.TE
.nr PS 9
.RT
.ad r
\fBTableau 2/I.340 [2T2.340], p.8\fR
.sp 1P
.RT
.ad b
.RT
.LP
.bp
.ce
\fBH.T. [3T2.340]\fR
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(342p) .
{
TABLE\ 2/I.340\ \fI(end)\fR
}
.TE
.TS
center box;
cw(36p) | cw(171p) sw(135p) , ^ | c | c.
ISDN connection type identity Attributes
{
Dominant attributes which define ISDN connection types
} Additional attributes
_
.TE
.TS
center box;
cw(12p) | cw(24p) | cw(24p) | cw(18p) | cw(36p) | cw(27p) | cw(36p) | cw(30p) | cw(18p) | cw(24p) | cw(15p) | cw(21p) | cw(21p) | cw(21p) | cw(15p) , ^ | ^ | l | l | l | l | l | l | l | l | l | l | l | l | l.
provide descriptions of Quality of Service and Network
Performance;
.LP
\(em
illustrate how the Quality of Service and the Network
Performance concepts are applied in digital networks,
including ISDNs;
.LP
\(em
describe the features of, and the relationship between, these
concepts;
.LP
\(em
indicate and classify performance concerns for which
parameters may be needed;
.LP
\(em
identify generic performance parameters.
.PP
The generic term \*Qperformance\*U refers to Quality of Service and to
Network Performance as they are defined in\ \(sc\ 1.2.
.sp 2P
.LP
1.2
\fIDescriptions of\fR \fIQuality of Service (QOS)\fR \fIand\fR
\fINetwork Performance (NP)\fR
.sp 1P
.RT
.sp 1P
.LP
1.2.1
\fIDescription of Quality of Service\fR
.sp 9p
.RT
.PP
QOS is defined in Recommendation G.106 \fI(Red Book)\fR | as follows:
\*QCollective effect of service performances which determine the degree of
satisfaction of a user of the service\*U.
.PP
The note of Recommendation \fI(Red Book)\fR | underlines that the QOS is
characterized by the combined aspects of:
.RT
.LP
\(em
service support and service operability performance,
and
.LP
\(em
servability and service integrity performance.
.PP
The definition of Quality of Service in Recommendation\ G.106
\fI(Red Book)\fR | is a wide one encompassing many areas of work, including
subjective customer satisfaction. However, within this Recommendation the
aspects of Quality of Service that are covered are restricted to the
identification of parameters that can be directly observed and measured
at the point at which the service is accessed by the user. Other types
of QOS
parameters which are subjective in nature, i.e.\ depend upon user actions or
subjective opinions, will not be specified in the I\(hySeries Recommendations
on QOS.
.bp
.sp 1P
.LP
1.2.2
\fIDescription of Network Performance\fR
.sp 9p
.RT
.PP
Network Performance is a statement of the performance of the
connection element or concatenation of connection elements employed to
provide a service. It is defined and measured in terms of parameters which
are
meaningful to the network provider and are used for the purposes of system
design, configuration, operation and maintenance. NP is defined independently
of terminal performance and user actions.
.PP
\fBnetwork performance (NP)\fR is defined as the ability of a network or
network portion to provide the functions related to communications between
users.
.PP
\fINote\fR \ \(em\ The performance of a network and its component parts
contributes to servability performance and service integrity performance as
defined in Recommendation\ G.106 \fI(Red Book)\fR , and is characterized
by a set of measurable and calculable parameters.
.RT
.sp 2P
.LP
\fB2\fR \fBPurpose of QOS and NP\fR
.sp 1P
.RT
.sp 1P
.LP
2.1
\fIGeneral\fR
.sp 9p
.RT
.PP
Bearer services and teleservices as described in the I.200\(hySeries Recommendations
are the objects which network and service providers offer to
their customers. A major attribute of these services is the set of QOS
parameters which a particular service offers. These parameters are user
oriented and take into account the elements involved in a particular service
as given in Figure\ 2/I.211.
.PP
Bearer services and teleservices are supported by a range of
connection types, each of which comprises several connection elements. The
performance of the connection types is characterized by a set of NP parameters.
These parameters are network oriented.
.PP
Figure\ 1/I.350 illustrates how the concepts of QOS and NP are applied
in the ISDN environment.
.RT
.LP
.rs
.sp 25P
.ad r
\fBFigure 1/I.350, (N), p.\fR
.sp 1P
.RT
.ad b
.RT
.LP
.bp
.sp 1P
.LP
2.2
\fIPurpose of QOS\fR
.sp 9p
.RT
.PP
A typical user is not concerned with how a particular service is
provided, or with any of the aspects of the network's internal design.
However, he is interested in comparing one service with another in terms
of certain
universal, user\(hyoriented performance concerns which apply to any end\(hyto\(hyend
service. Therefore from a user's point of view, Quality of Service is best
expressed by parameters which:
.RT
.LP
\(em
focus on user\(hyperceivable effects, rather than their causes
within the network;
.LP
\(em
do not depend, in their definition, on assumptions about the
network internal design;
.LP
\(em
take into account all aspects of the service from the user's
point of view which can be objectively measured at the service
access point;
.LP
\(em
may be assured to a user at the service access point by the
service provider(s);
.LP
\(em
are described in network independent terms and create a
common language understandable by both the user and the service
provider.
.sp 1P
.LP
2.3
\fIPurpose of NP\fR
.sp 9p
.RT
.PP
A network provider is concerned with the efficiency and
effectiveness of the network, in providing services to customers. Therefore
from a network provider's point of view, NP is best expressed by parameters
which provide information for:
.RT
.LP
\(em
system development;
.LP
\(em
network planning, both nationally and
internationally;
.LP
\(em
operation and maintenance.
.LP
\fB3\fR \fBPrinciples for the development of QOS and NP parameters and\fR
\fBvalues\fR
.sp 1P
.RT
.sp 2P
.LP
3.1
\fIGeneral principles\fR
.sp 1P
.RT
.sp 1P
.LP
3.1.1
\fIDistinction between QOS and NP\fR
.sp 9p
.RT
.PP
The user oriented QOS parameters provide a valuable framework for network
design, but they are not necessarily usable in specifying performance requirements
for particular connections. Similarly, the NP parameters
ultimately determine the (user observed) QOS, but they do not necessarily
describe that quality in a way that is meaningful to users. Both types of
parameters are needed, and their values must be quantitatively related if a
network is to be effective in serving its users. The definition of QOS
and NP parameters should make mapping of values clear in cases where there
is not a
simple one\(hyto\(hyone relationship between them.
.PP
Table 1/I.350 shows some of the characteristics which distinguish QOS and NP.
.RT
.sp 1P
.LP
3.1.2
\fIMeasurability of QOS and NP parameter values\fR
.sp 9p
.RT
.PP
Due to separating QOS and NP, a number of general points should be noted
when considering the development of parameters:
.RT
.LP
\(em
the definition of QOS parameters should be clearly based on
events and states observable at service access points and
independent of the network processes and events which support
the service;
.LP
\(em
the definition of NP parameters should be clearly based on
events and states observable at connection element boundaries,
e.g.\ protocol specific interface signals;
.LP
\(em
the use of events and states in the definition of parameters
should provide for measurements at the boundaries identified
above. Such measurements should be verifiable in accordance with
generally accepted statistical techniques.
.bp
.ce
\fBH.T. [T1.350]\fR
.ce
TABLE\ 1/I.350
.ce
\fBDistinction between quality of service and network performance\fR
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(60p) | cw(90p) .
Quality of service Network performance
_
.T&
lw(60p) | lw(90p) .
User oriented Provider oriented
_
.T&
lw(60p) | lw(90p) .
Service attribute Connection element attribute
_
.T&
lw(60p) | lw(90p) .
{
Focus on user\(hyobservable effects
} {
Focus on planning, development (design), operations and
maintenance
}
_
.T&
lw(60p) | lw(90p) .
{
Between (at) service access points
} {
End\(hyto\(hyend or network connection elements capabilities
}
_
.TE
.nr PS 9
.RT
.ad r
\fBTableau 1/I.350 [T1.350], p.\fR
.sp 1P
.RT
.ad b
.RT
.LP
.sp 3
.sp 1P
.LP
3.1.3
\fIMultiple network provider environments\fR
.sp 9p
.RT
.PP
It should be recognized in the development of parameter values that services
may be provided by multiple providers. In such an environment
different levels of QOS may be supported. Therefore, in practice, users may
experience a variety of ranges of QOS. It is thus important to establish
minimum performance levels for each service and for connection elements
providing international connections.
.RT
.sp 1P
.LP
3.2
\fIQOS principles\fR
.sp 9p
.RT
.PP
For the definition of parameters for QOS in the ISDN, the concept of bearer
services and teleservices needs to be borne in mind. In particular, there
is a difference between the kinds of parameters which would describe the
QOS of a bearer service and that of a teleservice, since the point of
observation of, or access to, the service is different in each case.
Figure\ 1/I.350 illustrates this point.
.PP
In the case of teleservices the interface between the user and the
service provider may be a man\(hymachine interface. In the case of bearer
services this interface corresponds to the S/T reference points. As a result,
some of
the parameters for describing the QOS of a teleservice will be different
from those which describe the QOS of a bearer service.
.PP
In describing the QOS of teleservicess, the performance of the
terminal equipment (TE) has to be taken into account. For a teleservice,
there should be a mapping between the QOS of the teleservice and the performance
of the customer equipment including the terminal and the overall (end\(hyto\(hyend)
NP of the connection elements supporting this service.
.PP
For bearer service there should be mapping between the QOS of the
bearer service and the overall (end\(hyto\(hyend) NP of connection elements
supporting this service.
.RT
.sp 1P
.LP
3.3
\fINP principles\fR
.sp 9p
.RT
.PP
When developing NP parameters the following points should be borne in mind:
.RT
.LP
\(em
NP parameters must be measurable at the boundary of the
network connection element(s) to which they are applied. The
definitions should not be based on assumptions about either the
internal characteristics of a network (or portions thereof),
or the internal causes of impairments observed at the
boundaries;
.bp
.LP
\(em
the division of a network portion into sub\(hycomponents
should only be done if they must be specified separately in
order to ensure satisfactory end\(hyto\(hyend performance or, where
appropriate, to derive fair and reasonable allocations among
providers. No network provider should bear a disproportionate
cost in establishing and operating a service.
.sp 2P
.LP
3.4
\fIPrimary and derived performance parameters\fR
.sp 1P
.RT
.sp 1P
.LP
3.4.1
\fIDescription\fR
.sp 9p
.RT
.sp 1P
.LP
\fBprimary performance parameter\fR
.sp 9p
.RT
.PP
A parameter or a measure of a parameter determined on the basis of direct
observations of events at services access points or connection element
boundaries.
.RT
.sp 1P
.LP
\fBderived performance parameter\fR
.sp 9p
.RT
.PP
A parameter or a measure of a parameter determined on the basis of observed
values of one or more relevant primary performance parameters and
decision thresholds for each relevant primary performance
parameter.
.RT
.sp 1P
.LP
3.4.2
\fIRelationship between primary and derived performance parameters\fR
.sp 9p
.RT
.PP
A number of event types can be directly observed at service access points
or connection element boundaries. Examples of such events
are:
.RT
.LP
\(em
the layer 3 protocol state transition associated with the
transfer of a SETUP message or a DISCONNECT message across a
connection element boundary;
.LP
\(em
the correct receipt of an information bit (or a specified
number of information bits) at an interface.
.PP
Parameters related to the time interval between specific events
and the frequency of events can be measured. These directly measurable
parameters or primary performance parameters describe the QOS (at service
access points) or the NP (at connection element boundaries) during periods
when the service or connection is available.
.PP
Derived performance parameters describe performance based on events
which are defined as occurring when the value of a function of a primary
performance parameter(s) crosses a particular threshold. These derived
threshold events identify the transitions between the available and the
unavailable states. Parameters related to the time interval between these
derived threshold events and their frequency can be identified. These derived
performance parameters describe the QOS and the NP for all time intervals;
i.e.\ during periods when the service or connection is available or
unavailable.
.PP
\fINote\fR \ \(em\ Primary performance parameters are measured for all time
intervals, since the transitions between available and unavailable states
depend upon the value of these parameters. However, the values of primary
performance parameters would not be specified for a service or connection in
the unavailable state.
.RT
.sp 2P
.LP
\fB4\fR \fBGeneric performance parameters\fR
.sp 1P
.RT
.PP
Nine generic primary performance parameters are listed below. These have
been developed as a result of the matrix approach described in Annex\ A.
These parameters may be used in developing specific QOS and NP
parameters:
.RT
.LP
\(em
access speed;
.LP
\(em
access accuracy;
.LP
\(em
access dependability;
.LP
\(em
information transfer speed;
.LP
\(em
information transfer accuracy;
.bp
.LP
\(em
information transfer dependability;
.LP
\(em
disengagement speed;
.LP
\(em
disengagement accuracy;
.LP
\(em
disengagement dependability.
.PP
Section 3.4 defines derived performance parameters in addition to primary
parameters. Derived performance parameters are determined utilizing a function
of the primary performance parameter values. Recommendation\ G.821
defines one such function, which identifies transitions between available
and unavailable states based on a threshold for severely errored seconds.
The
generic derived performance parameter associated with such a function is
availability.
.PP
Examples of specific primary and derived performance parameters for
bearer service QOS and those for circuit\(hyswitched and packet\(hyswitched
NP are
provided in Annex\ B.
\v'6p'
.RT
.ce 1000
ANNEX\ A
.ce 0
.ce 1000
(to Recommendation I.350)
.sp 9p
.RT
.ce 0
.ce 1000
\fBMethod of identifying parameters\fR
.sp 1P
.RT
.ce 0
.LP
A.1
\fIThe matrix approach\fR
.sp 1P
.RT
.PP
The matrix provides a systematic method of identifying and
organizing candidate network performance parameters with the objective of
defining a concise set of parameters and, where appropriate, their QOS
counterparts. This tool should be used as the basis for collection and
evaluation of network performance parameters for digital networks, including
ISDNs.
.RT
.sp 2P
.LP
A.2
\fI3\ \(mu\ 3 matrix approach for network performance\fR
.sp 1P
.RT
.PP
The 3\ \(mu\ 3 matrix approach for network performance is illustrated in
Figure\ A\(hy1/I.350. The main features are as follows:
.RT
.LP
1)
Each row represents one of the three basic and distinct
communication functions.
.LP
\fINote\fR \ \(em\ The access function represents the connectionless
as well as connection oriented services which are possible with
ISDNs.
.LP
2)
Each column represents one of the three mutually exclusive
outcomes possible when a function is attempted.
.LP
3)
The 3\ \(mu\ 3 matrix parameters are defined on the basis of
events at connection element boundaries and are termed \*Qprimary
performance parameters\*U. \*QDerived performance parameters\*U are
defined on the basis of a functional relationship of primary
performance parameters, outage thresholds and an observation
interval.
.LP
4)
NP primary performance parameters should be defined so as to
be measurable at the boundaries of the connection element(s) to
which they apply. NP parameter definitions should not depend
upon assumptions about impairment causes that are not
detectable at the boundaries.
.LP
5)
Availability is a derived performance parameter. Decisions
on the appropriate primary performance parameters, outage
threshold and algorithms for its definition require further
detailed study.
.PP
\fINote\fR \ \(em\ The following terminology problems are pointed out.
Appropriate terms should be selected after further study:
.LP
a)
the term \*Qaccess\*U is used. However, the term \*Qselection (of
the connection type, the destination and facility)\*U has been
proposed as an alternative.
.LP
b)
The term \*Q
dependability
\*U is used. However, the
definition of dependability as used here is somewhat different
from that in Recommendation\ G.106 \fI(Red Book)\fR . Alternative
terms, \*Qinserveability\*U and \*Qrefusal\*U are proposed.
.LP
c)
The term \*Qavailability\*U is provisionally used. An
alternative term \*Qacceptability\*U has been proposed.
.bp
.LP
.rs
.sp 29P
.ad r
\fBFigure A\(hy1/I.350, (N), p.\fR
.sp 1P
.RT
.ad b
.RT
.sp 2P
.LP
A.3
\fI3\ \(mu\ 3 matrix approach for QOS\fR
.sp 1P
.RT
.PP
The same 3\ \(mu\ 3 matrix approach as that described for network
performance may be used for the related Quality of Service
parameters.
.PP
QOS parameters should be defined so as to be measurable at service
access points. QOS parameter definitions should not depend upon assumptions
of impairment causes that are not detectable at the service access
points.
.PP
Loss of service parameters are considered to be derived QOS
parameters. An alternative matrix has been proposed and is still under
consideration.
.RT
.sp 2P
.LP
A.4
\fIDescription of the basic communication functions\fR
.sp 1P
.RT
.sp 1P
.LP
A.4.1
\fIAccess\fR
.sp 9p
.RT
.PP
The access function begins upon issuance of an access request
signal or its implied equivalent at the interface between a user and the
communication network. It ends when either:
.RT
.LP
1)
a ready for data or equivalent signal is issued to the
calling users, or
.LP
2)
at least one bit of user information is input to the
network (after connection establishment in connection\(hyoriented
services).
.PP
It includes all activities traditionally associated with physical circuit
establishment (e.g.\ dialling, switching, and ringing) as well as any
activities performed at higher protocol layers.
.bp
.sp 1P
.LP
A.4.2
\fIUser information transfer\fR
.sp 9p
.RT
.PP
The user information transfer begins on completion of the access
function, and ends when the \*Qdisengagement request\*U terminating a communication
session is issued. It includes all formatting, transmission, storage, error
control and media conversion operations performed on the user information
during this period, including necessary retransmission within the
network.
.RT
.sp 1P
.LP
A.4.3
\fIDisengagement\fR
.sp 9p
.RT
.PP
There is a disengagement function associated with each participant in a
communication session: each disengagement function begins on issuance of
a disengagement request signal. The disengagement function ends, for each
user, when the network resources dedicated to that user's participation
in the
communication session have been released. Disengagement includes both physical
circuit disconnection (when required) and higher\(hylevel protocol termination
activities.
.RT
.sp 2P
.LP
A.5
\fIDescription of the performance\fR
.sp 1P
.RT
.sp 1P
.LP
A.5.1
\fISpeed\fR
.sp 9p
.RT
.PP
Speed is the performance criterion that describes the time interval that
is used to perform the function or the rate at which the function is
performed. (The function may or may not be performed with the desired
accuracy.)
.RT
.sp 1P
.LP
A.5.2
\fIAccuracy\fR
.sp 9p
.RT
.PP
Accuracy is the performance criterion that describes the degree of correctness
with which the function is performed. (The function may or may not be performed
with the desired speed.)
.RT
.sp 1P
.LP
A.5.3
\fIDependability\fR
.sp 9p
.RT
.PP
Dependability is the performance criterion that describes the
degree of certainty (or surety) with which the function is performed regardless
of speed or accuracy, but within a given observation interval.
\v'6p'
.RT
.ce 1000
ANNEX\ B
.ce 0
.ce 1000
(to Recommendation I.350)
.sp 9p
.RT
.ce 0
.ce 1000
\fBRelationship between generic and possible specific QOS and NP
parameters\fR
.sp 1P
.RT
.ce 0
.PP
This Annex illustrates the qualitative relationship between
the generic parameters defined in this Recommendation and a candidate set of
specific QOS and NP parameters. Tables\ B\(hy1/I.350, B\(hy2/I.350 and\
B\(hy3/I.350
illustrate the relationship between the generic parameters and specific
bearer service QOS, circuit\(hyswitched NP, and packet\(hyswitched NP parameters,
respectively.
.sp 1P
.RT
.LP
.rs
.sp 10P
.ad r
Blanc
.ad b
.RT
.LP
.bp
.ce
\fBH.T. [T2.350]\fR
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(342p) .
TABLE\ B\(hy1/I.350
.T&
cw(342p) .
{
\fBQualitative relationship between generic performance parameters
\fBand candidate bearer service QOS parameters\fR
}
.TE
.TS
center box;
lw(72p) | cw(18p) sw(18p) sw(18p) sw(18p) sw(18p) sw(18p) sw(18p) sw(18p) sw(18p) sw(18p) sw(18p) sw(18p) | cw(18p) sw(18p) sw(18p) , ^ | l | l | l | l | l | l | l | l | l | l | l | l | l | l | l.
\fBQualitative relationship between generic performance
parameters and candidate circuit\(hyswitched NP parameters\fR
}
.TE
.TS
center box;
lw(80p) | cw(20p) sw(14p) sw(20p) sw(20p) sw(20p) sw(20p) sw(20p) sw(14p) sw(20p) sw(20p) sw(20p) sw(20p) | cw(20p) sw(14p) , ^ | l | l | l | l | l | l | l | l | l | l | l | l | l | l.
\fBQualitative relationship between generic performance
parameters and candidate packet\(hyswitched NP parameters\fR
}
.TE
.TS
center box;
lw(80p) | cw(20p) sw(14p) sw(20p) sw(20p) sw(20p) sw(20p) sw(20p) sw(14p) sw(20p) sw(20p) sw(20p) sw(20p) | cw(20p) sw(14p) , ^ | l | l | l | l | l | l | l | l | l | l | l | l | l | l.
{
Packet\(hyswitched
NP parameters
Generic parameters
} {
Primary performance parameters
} {
Derived performance parameters
Virtual circuit set\(hyup delay
Virtual circuit set\(hyup error probability
Virtual circuit set\(hyup denial probability
Date packet transfer delay
Throughput capacity
Residual error rate
Reset probability
Reset stimulus probability
Virtual circuit clearing delay
Virtual circuit clearing denial probability
Virtual circuit premature disconnect probability
Virtual circuit premature disconnect stimulus probability
The connection processing control states have been divided into
those used in incoming and outgoing circuit handling. The usage of the term
direction in this contex refers to the direction of the connection
.nr PS 9
.RT
.ad r
\fBTableau 2/I.352 [T2.352], p.21\fR
.sp 1P
.RT
.ad b
.RT
.LP
.rs
.sp 16P
.ad r
Blanc
.ad b
.RT
.LP
.bp
.LP
.rs
.sp 25P
.ad r
\fBFigure 3/I.352, (N), p.30\fR
.sp 1P
.RT
.ad b
.RT
.ce
\fBH.T. [T3.352]\fR
.ce
TABLE\ 3/I.352
.ce
\fBMessage transfer events for measuring connection set\(hyup delay\fR
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(72p) | cw(48p) sw(48p) , ^ | c | c.
Connection element boundary Message transfer event
Starting event number Ending event number
_
.T&
lw(72p) | cw(48p) | cw(48p) .
Calling S/T interface {
1 (en bloc) or
5 (overlap sending)
} 11
_
.T&
lw(72p) | cw(48p) | cw(48p) .
Called S/T interface \fBS\fR 2 10
_
.T&
lw(72p) | cw(48p) | cw(48p) .
{
Access/national transit (originating)
} S1 S5
_
.T&
lw(72p) | cw(48p) | cw(48p) .
{
Access/national transit (terminating)
} S2 S6
_
.T&
lw(72p) | cw(48p) | cw(48p) .
{
National/international transit (originating)
} S2 S6
_
.T&
lw(72p) | cw(48p) | cw(48p) .
{
National/international transit (terminating)
} S1 S5
.TE
.LP
\fINote\fR
\ \(em\ En bloc and overlap sending options at the calling S/T
interface.
.nr PS 9
.RT
.ad r
\fBTableau 3/I.352 [T3.352], p.22\fR
.sp 1P
.RT
.ad b
.RT
.LP
.bp
.ce
\fBH.T. [T4.352]\fR
.ce
TABLE\ 4/I.352
.ce
\fBOverall connection set\(hyup delay\fR
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(54p) | cw(42p) | cw(42p) .
ISDN connection type Statistic Connection set\(hyup delay
_
.T&
lw(54p) | cw(42p) | cw(42p) , ^ | c | c.
{
No. 1: 64 kbit/s unrestricted switched
} Mean 4500 ms | ua\d\u)\d
95% 8350 ms\ua\d\u)\d
.TE
.LP
\ua\d\u)\d
Provisional values; the actuel target values are for further
study.
.LP
\fINote\ 1\fR
\ \(em\ The values take into account worst case situations such as the
longest length reference connection (27\ 500\ km) as specified in
Recommendation\ G.104.
.LP
The values observed will be dominated by the number of exchanges in a
connection. For the moderate length reference connection (11\ 000\ km), the
observed values will be lower.
.LP
\fINote\ 2\fR
\ \(em\ Delays are specified for a nominal busy hour.
.LP
\fINote\ 3\fR
\ \(em\ Connection set\(hyup attempts which exceed a specified timeout
value are excluded in computing these statistics and are counted separately as connection set\(hyup denials.
.LP
\fINote\ 4\fR
\ \(em\ In this table, the relevant ISDN connection types given in
Table\ 2/I.340 are specified.
.LP
\fINote\ 5\fR
\ \(em\ Those message processing delays that are dependent on a user
equipment network are not included. In addition, when transmitting a signal
message defined in Recommendation\ Q.931 from the network to a user, before the message actually passes across the S/T reference point, it may have to wait in the exchange or signalling system while another message (signal or user packet) is being transmitted to the user. Since this waiting time depends on the volume of user packet (message) traffic over the D\(hychannel, the resulting delay is
beyond the responsibility of the network provider.
.LP
\fINote\ 6\fR
\ \(em\ The values take into account the additional signalling points for
the 95% case of the hypothetical signalling reference connection in
Recommendation\ Q.709.
.LP
\fINote\ 7\fR
\ \(em\ The delay objectives in the table are primarily applicable to
connections provided exclusively over ISDNs, i.e. no interworking.
.LP
\fINote\ 8\fR
\ \(em\ The connection set\(hyup and disconnected procedures in ISDNs for
circuit\(hymode voice and data are essentially the same; therefore, the delay
definitions are applicable for circuit\(hymode voice and circuit\(hymode data. The
provisional values in the tables are applicable for both circuit\(hymode voice and circuit\(hymode data with no interworking. However, the observed delay performance may not be identical due to network architectural differences and
interworking.
.nr PS 9
.RT
.ad r
\fBTableau 4/I.352 [T4.352], p.23\fR
.sp 1P
.RT
.ad b
.RT
.ad r
.rs
.sp 5P
Blanc
.ad b
.RT
.LP
.bp
.ce
\fBH.T. [T5.352]\fR
.ce
TABLEAU\ 5/I.352
.ce
\fBMessage transfer events for measuring alerting delay\fR
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(72p) | cw(48p) sw(48p) , ^ | c | c.
Connection element boundary Message transfer event
Starting event number Ending event number
_
.T&
lw(72p) | cw(48p) | cw(48p) .
Calling S/T interface {
1 (en bloc) or
5 (overlap sending)
} \fBS\fR 9
_
.T&
lw(72p) | cw(48p) | cw(48p) .
Called S/T interface \fBS\fR 2 \fBS\fR 6
_
.T&
lw(72p) | cw(48p) | cw(48p) .
{
Access/national transit (originating)
} S1 S3
_
.T&
lw(72p) | cw(48p) | cw(48p) .
{
Access/national transit (terminating)
} S2 S4
_
.T&
lw(72p) | cw(48p) | cw(48p) .
{
National/international transit (originating)
} S2 S4
_
.T&
lw(72p) | cw(48p) | cw(48p) .
{
National/international transit (terminating)
} S1 S3
.TE
.LP
\fINote\fR
\ \(em\ En bloc and overlap sending options at the calling S/T
interface.
.nr PS 9
.RT
.ad r
\fBTableau 5/I.352 [T5.352], p.24\fR
.sp 1P
.RT
.ad b
.RT
.LP
.sp 4
.sp 1P
.LP
3.1.2.2
\fIDefinition of alerting delay between two connection\fR
\fIelement boundaries\fR
.sp 9p
.RT
.PP
The \fBalerting delay between two connection element boundaries\fR can
be measured at one connection element boundary, B\di\u, and then measured at
another boundary, B\dj\u, further from the calling S/T interface. The difference
in the values obtained is the alerting delay contributed by the connection
elements between the two boundaries.
.PP
Alerting delay between two connection element boundaries =
\fId\fR \fI\fI\d\fIi\fR\uis the alerting delay measured at B\di\u,
.LP
\fId\fR \fI\fI\d\fIj\fR\uis the alerting delay measured at B\dj\u.
.PP
The overall alerting delay is the alerting delay between the two S/T interfaces,
B\d1\uand B\dn\uin Figure\ 1/I.352 for the reference
configuration types in Recommendation\ I.340. This overall alerting delay
excludes the called user response time. The alerting delay for a connection
element is the alerting delay between the boundaries delimiting that
connection element.
.sp 1P
.LP
3.1.2.3
\fIAlerting delay specification\fR
.sp 9p
.RT
.PP
The overall alerting delay should not exceed the values given in
Table\ 6/I.352.
.PP
The allocation of the alerting delay among the elements of the
connections are for further study.
.bp
.RT
.ce
\fBH.T. [T6.352]\fR
.ce
TABLE\ 6/I.352
.ce
\fBOverall alerting delay\fR
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(54p) | cw(42p) | cw(42p) .
ISDN connection type Statistic Alerting delay
_
.T&
lw(54p) | cw(42p) | cw(42p) , ^ | c | c.
{
No. 1: 64 kbit/s unrestricted switched
} Mean 4500 ms | ua\d\u)\d
95% 8350 ms\ua\d\u)\d
.TE
.LP
\ua\d\u)\d
Provisional values; the actual target values are for further
study.
.LP
\fINote\ 1\fR
\ \(em\ The values take into account worst case situations such as the
longest length reference connection (27\ 500\ km) as specified in
Recommendation\ G.104.
.LP
The values observed will be dominated by the number of exchanges in a
connection. For the moderate length reference connection (11\ 000\ km) the
observed values will be lower.
.LP
\fINote\ 2\fR
\ \(em\ Delays are specified for a nominal busy hour.
.LP
\fINote\ 3\fR
\ \(em\ Connection set\(hyup attempts which exceed a specified timeout value
are excluded in computing these statistics and are counted separately as
connection set\(hyup denials.
.LP
\fINote\ 4\fR
\ \(em\ In this table the relevant ISDN connection types given in
Table\ 2/I.340 are specified.
.LP
\fINote\ 5\fR
\ \(em\ Those message processing delays that are dependent on a user
equipment/network are not included. In addition, when transmitting a signal
message defined in Recommendation\ Q.931 from the network to a user, before the message actually passes across the S/T reference\(hypoint, it may have to wait in the exchange or signalling system while another message (signal or user packet) is being transmitted to the user. Since this waiting time depends on the volume of user packet (message) traffic over the D\(hychannel, the resulting delay is
beyond the responsibility of the network provider.
.LP
\fINote\ 6\fR
\ \(em\ The values take into account the additional signalling points for
the 95% case of the hypothetical signalling reference connection in
Recommendation\ Q.709.
.LP
\fINote\ 7\fR
\ \(em\ The delay objectives in the table are primarily applicable to
connections provided exclusively over ISDNs, i.e. no interworking.
.LP
\fINote\ 8\fR
\ \(em\ The connection set\(hyup and disconnect procedures in ISDNs for
circuit\(hymode voice and data are essentially the same. Therefore, the delay
definitions are applicable for circuit\(hymode voice and circuit\(hymode data. The
provisional values in the tables are applicable for both circuit\(hymode voice and circuit\(hymode data with no interworking. However, the observed delay performance may not be identical due to network architectural differences and
interworking.
.nr PS 9
.RT
.ad r
\fBTableau 6/I.352 [T6.352], p.\fR
.sp 1P
.RT
.ad b
.RT
.LP
.sp 2
.sp 2P
.LP
3.2
\fIDisconnect phase parameters\fR
.sp 1P
.RT
.sp 1P
.LP
3.2.1
\fIDisconnect delay\fR
.sp 9p
.RT
.PP
Disconnect definition is based only on a one\(hyway message transport from
the clearing party to be cleared party. Therefore, this parameter requires
observations at two connection element boundaries.
.bp
.RT
.sp 1P
.LP
3.2.1.1
\fIDefinition of disconnect delay between two connection element\fR \fIboundaries\fR
.sp 9p
.RT
.PP
\fBDisconnect delay between two connection element boundaries,\fR
\fBB\fR\(da\fBi and B\fR\(da\fBj\fR , is defined as the length of time
that starts when a
DISConnect message creates a message transfer event at B\di\uand ends when
that DISConnect message creates a message transfer event at B\dj\u, further
from the clearing party S/T interface.
.PP
Disconnect delay between two connection element boundaries =
(\fIt\fR\d2\u\ \(em\ \fIt\fR\d1\u)
.RT
.LP
where
.LP
\fIt\fR\d1\uis the time of occurrence for the message transfer
event at B\di\u,
.LP
\fIt\fR\d2\uis the time of occurrence for the message transfer
event at B\dj\u.
.PP
The overall disconnect delay is the disconnect delay between two S/T interfaces,
B\d1\uand B\dn\uin Figure\ 1/I.352 for the reference
configuration types in Recommendation\ I.340. The disconnect delay for a
connection element is the disconnect delay between the boundaries delimiting
that connection element. The specific message transfer events used in measuring
disconnect delay are shown in Table\ 7/I.352.
.LP
.sp 1
.ce
\fBH.T. [T7.352]\fR
.ce
TABLE\ 7/I.352
.ce
\fBMessage transfer events for measuring disconnect delay\fR
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(72p) | cw(72p) sw(72p) , ^ | c | c.
Connection element(s) {
Message transfer event
(at connection element boundary)
}
Starting event number Ending event number
_
.T&
lw(72p) | lw(72p) | lw(72p) .
S/T to S/T interface 14 (Clearing end) 15 (Cleared end)
_
.T&
lw(72p) | lw(72p) | lw(72p) .
National transit S7 (Access/national transit) {
S8 (National/international transit)
}
_
.T&
lw(72p) | lw(72p) | lw(72p) .
International transit {
S8 (National/international transit)
} {
S7 (International/national transit)
}
_
.TE
.nr PS 9
.RT
.ad r
\fBTableau 7/I.352 [T7.352], p.\fR
.sp 1P
.RT
.ad b
.RT
.LP
.sp 1
.sp 1P
.LP
3.2.1.2
\fIDisconnect delay specification\fR
.sp 9p
.RT
.PP
The overall disconnect delay should not exceed the values given in Table\
8/I.352.
.PP
The disconnect delay values for connection elements are for further
study.
.RT
.sp 1P
.LP
3.2.2
\fIRelease delay\fR
.sp 9p
.RT
.PP
Release delay is defined only at the clearing party S/T
interface.
.RT
.sp 1P
.LP
3.2.2.1
\fIDefinition of release delay\fR
.sp 9p
.RT
.PP
\fBrelease delay\fR is defined as the length of time that starts
when a DISConnect message from the clearing party creates a message transfer
event at the clearing party S/T interface and ends when the RELease message
creates a message transfer event at the same interface.
.PP
Release delay at the clearing part S/T interface =
(\fIt\fR\d2\u\ \(em\ \fIt\fR\d1\u)
.RT
.LP
where
.LP
\fIt\fR\d1\uis the time of occurrence for the starting message
transfer event,
.LP
\fIt\fR\d2\uis the time of occurrence for the ending message
transfer event.
.PP
Since the release message sent by the exchange at the clearing end is only
transported over the access connection element at that end, the
distinction between overall delay and connection element delay is not relevant.
The specific message transfer events used in measuring release delay are
shown in Table\ 9/I.352.
.bp
.ce
\fBH.T. [T8.352]\fR
.ce
TABLE\ 8/I.352
.ce
\fBDisconnect delay\fR
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(54p) | cw(42p) | cw(42p) .
ISDN connection type Statistic Disconnect delay
_
.T&
lw(54p) | cw(42p) | cw(42p) , ^ | c | c.
{
No. 1: 64 kbit/s unrestricted switched
} Mean 2700 ms\ua\d\u)\d
95 % 4700 ms\ua\d\u)\d
.TE
.LP
\ua\d\u)\d
Provisional values; the actual target values are for further
study.
.LP
\fINote\ 1\fR
\ \(em\ The values take into account worst case situation such as the
longest length reference connection (27\ 500\ km) as specified in
Recommendation\ G.104.
.LP
The values observed will be dominated by the number of exchanges in a
connection. For the moderate length reference connection (11\ 000\ km) the
observed values will be lower.
.LP
\fINote\ 2\fR
\ \(em\ Delays are specified for a nominal busy hour.
.LP
\fINote\ 3\fR
\ \(em\ In this table the relevant ISDN connection types given in
Table\ 2/I.340 are specified.
.LP
\fINote\ 4\fR
\ \(em\ The values take into account the additional signalling points for
the 95% case of the hypothetical signalling reference connection in
Recommendation\ Q.709.
.LP
\fINote\ 5\fR
\ \(em\ The delay objectives in the table are primarily applicable to
connections provided exclusively over ISDNs, i.e. no interworking.
.LP
\fINote\ 6\fR
\ \(em\ The connection set\(hyup and disconnect procedures in ISDNs for
circuit\(hymode voice and data are essentially the same. Therefore, the delay
definitions are applicable for circuit\(hymode voice and circuit\(hymode data. The
provisional values in the tables are applicable for both circuit\(hymode voice
and circuit\(hymode data with no interworking. However, the observed delay
performance may not be identical due to network architectural differences and interworking.
.nr PS 9
.RT
.ad r
\fBTableau 8/I.352 [T8.352], p.27\fR
.sp 1P
.RT
.ad b
.RT
.LP
.rs
.sp 16P
.ad r
Blanc
.ad b
.RT
.LP
.bp
.ce
\fBH.T. [T9.352]\fR
.ce
TABLE\ 9/I.352
.ce
\fBMessage transfer events for measuring release delay\fR
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(72p) | cw(48p) sw(48p) , ^ | c | c.
Connection element boundary Message transfer event
Starting event number Ending event number
_
.T&
lw(72p) | lw(48p) | lw(48p) .
Clearing party S/T 14 16
_
.T&
lw(72p) | lw(48p) | lw(48p) .
Cleared party S/T Not applicable Not applicable
_
.T&
lw(72p) | lw(48p) | lw(48p) .
Access/National transit Not applicable Not applicable
_
.T&
lw(72p) | lw(48p) | lw(48p) .
{
National/international transit
} Not applicable Not applicable
_
.TE
.nr PS 9
.RT
.ad r
\fBTableau 9/I.352 [T9.352], p.28\fR
.sp 1P
.RT
.ad b
.RT
.LP
.sp 3
.sp 1P
.LP
3.2.2.2
\fIRelease delay specification\fR
.sp 9p
.RT
.PP
The release delay should not exceed the values given in
Table\ 10/I.352.
.RT
.LP
.sp 3
.ce
\fBH.T. [T10.352]\fR
.ce
TABLE\ 10/I.352
.ce
\fBRelease delay\fR
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(54p) | cw(42p) | cw(42p) .
ISDN connection type Statistic Release delay
_
.T&
lw(54p) | cw(42p) | cw(42p) , ^ | c | c.
{
No. 1: 64 kbit/s unrestricted switched
} Mean 300 ms | ua\d\u)\d
95% 850 ms | ua\d\u)\d
.TE
.LP
\ua\d\u)\d
Provisional values; the actual target values are for further
study.
.LP
\fINote\ 1\fR
\ \(em\ The delay objectives in the table are primarily applicable to
connections provided exclusively over ISDNs, i.e. no interworking.
.LP
\fINote\ 2\fR
\ \(em\ The connection set\(hyup and disconnect procedures in ISDNs for
circuit\(hymode voice and data are essentially the same. Therefore, the delay
definitions are applicable for circuit\(hymode voice and circuit\(hymode data. The provisional values in the tables are applicable for both circuit\(hymode voice and circuit\(hymode data with no interworking. However, the observed delay performance may not be identical due to network architectural differences and