home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Internet Standards
/
CD2.mdf
/
ccitt
/
1992
/
e
/
e412.asc
< prev
next >
Wrap
Text File
|
1991-12-30
|
29KB
|
568 lines
All drawings appearing in this Recommendation have been done in Autocad.
Recommendation E.412
NETWORK MANAGEMENT CONTROLS
1 Introduction
1.1 Network management controls provide the means to alter the flow of traffic
in the network in support of the network management objectives given in
Recommendation E.410. Most network management controls are taken by or in the
exchange (see Recommendation Q.542), but certain actions can be taken external to
the exchange. This Recommendation provides specific information on network
management controls and gives guidance concerning their application. However, it
should be noted that the suggested use for each network management control is
given only for the purpose of illustration. Other controls, separately or in
combination, may be more appropriate in any given situation.
1.2 The application or removal of network management controls should be based
on network performance data which indicates that action is required in accordance
with the network management principles in Recommendation E.410, S 4. Performance
data will also measure the effect of any network management control taken, and
will indicate when a network management control should be modified or removed
(see Recommendations E.411 and E.502).
1.3 Controls can be activated or removed in an exchange by input from a
network management operations system or by direct input from a terminal. In some
cases, controls can be activated automatically either by external or internal
stimulus, or when a parameter threshold has been exceeded. [The automatic
congestion control (ACC) system is an example (see S 4.1).] When automatic
control operation is provided, means for human override should also be provided.
2 Traffic to be controlled
2.1 Type of traffic
Exchanges should be capable of applying a range of network management
controls (see Recommendation Q.542). For increased flexibility and precision,
there is considerable advantage when the effect of a control can be limited to a
particular specified traffic element.
The operating parameters of a control can be defined by a set of traffic
attributes. As shown in Figure 1/E.412, these parameters include distinctions
based on the origin of the traffic, for example customer-dialled,
operator-dialled, transit or other such classification as may be specified by the
Administration. These can be further classified by type of service, particularly
for ISDN.
Figure 1/E.412 - T0200760-87
Additional attributes can be specified based on information which may be
available in the exchange. For example, incoming/outgoing circuit group class, or
hard-to-reach status of destinations (see S 2.2) can be used. Further
distinctions can be based on the outgoing traffic type, for example direct
routed, alternate routed or transit.
In general, the more attributes that can be specified for a control, the
more precise will be its effect.
Note - Precision is of vital importance, particularly in the case of
protective controls.
2.2 Hard-to-reach (HTR) process
2.2.1 A hard-to-reach process for network management will enable exchanges to
automatically make more efficient use of network resources during periods of
network congestion by improving the performance of network management controls.
This improved performance is derived from the ability to distinguish between
destinations that are easy to reach (ETR) and destinations that are hard-to-reach
(HTR), (e.g., destinations with a low answer bid ratio) and applying heavier
controls to HTR traffic. This distinction can be based on:
i) internal performance measurements within the exchange and/or the
network management operations system;
ii) similar information gathered and reported by other exchanges;
iii) historical and current observations of network performance by
network managers.
The network manager should have the ability to set the threshold for HTR
determination in the exchange or network management operations system, and to
assign a destination as HTR regardless of its actual status.
2.2.2 Controlling traffic based on HTR status
Fascicle II.3 - Rec. E.412 PAGE1
When a call to a destination that is on the HTR list is being routed and a
network management control on HTR traffic is encountered, the call should be
controlled according to the relevant parameters. If a destination is considered
HTR, it normally should be HTR for all outgoing circuit groups.
Additional details of the hard-to-reach process can be found in
Recommendation Q.542.
2.3 Methods for specifying the amount of traffic to be controlled
2.3.1 Call percentage control
There is considerable advantage when exchange controls can be activated to
affect a variable percentage of traffic (for example 10%, 25%, 50%, 75% or 100%).
2.3.2 Call rate control
An ability to set an upper limit on the maximum number of calls to be
allowed to access the network during a specified period of time is of particular
advantage.
3 Exchange controls
Network management controls may be applied in exchanges to control traffic
volume or to control the routing of traffic. The resulting effect on traffic of
these controls may be expansive or protective, depending on the control used, its
point of application and the traffic element selected for control.
3.1 Traffic volume controls
Traffic volume controls generally serve to control the volume of traffic
offered to a circuit group or destination. These include the following:
3.1.1 Destination controls
3.1.1.1 Code blocking
This control bars routing for a specific destination on a percentage
basis. Code blocking can be done on a country code, an area code, an exchange
identifying code or an individual line number. The last of these is the most
selective control available.
Typical application: Used for immediate control of focussed overloads or
mass-calling situations.
3.1.1.2 Call-gapping
This control sets an upper limit on the number of call attempts allowed to
be routed to the specified destination in a particular period of time (for
example, no more than 5 call attempts per minute). Thus, the number of call
attempts that are routed can never exceed the specified amount.
Typical application: Used for the control of focussed overloads,
particularly mass-calling to an individual line number. A detailed
analysis may be required to determine the proper call-rate parameters.
3.1.2 Cancellation of direct routing
This control blocks the amount of direct routed traffic accessing a
circuit group.
Typical application: Used to reduce traffic to congested circuit groups or
exchanges where there is no alternate routed traffic.
3.1.3 Circuit directionalization
This control changes both-way operated circuits to incoming operated
circuits, either on a percentage basis or by a specified number of circuits. At
the end of the circuit group for which access is inhibited, this is a protective
action, whereas at the other end of the circuit group (where access is still
available), it is an expansive action.
Typical application: To enhance the flow of traffic outward from a
disaster area while inhibiting incoming traffic. To have an effect, it is
recommended that the minimum amount of directionalization be at least 50%.
3.1.4 Circuit turndown/busying/blocking
This control removes one-way and/or both-way operated circuits from
service, either on a percentage basis or by a specified number of circuits.
Typical application: Used to control exchange congestion when no other
control action is available.
3.1.5 Specialized volume controls
Both the automatic congestion control (ACC) system and the selective
circuit reservation control (SCR) are volume controls, but due to their
specialized nature, they are described separately in S 4.1 and S 4.2.
3.2 Routing control
Routing controls are used to control the routing of traffic to a
destination, or to or from a circuit group. However, it should be noted that in
some cases a routing control may also affect the volume of traffic. Controls
PAGE2 Fascicle II.3 - Rec. E.412
which are applied to circuit groups may also be applied to circuit sub-groups,
when appropriate.
3.2.1 Cancellation of alternative routing
Two versions of this control are possible. One version prevents traffic
from overflowing FROM the controlled circuit group: alternative routing from
(ARF). The other version prevents overflow traffic from all sources from having
access TO the controlled circuit group: alternative routing to (ART). See Figure
2/E.412.
Typical application: There are many uses for this control. These include
controlling alternative routing in a congested network to limit multi-link
connections, or to reduce alternative routed attempts on a congested
exchange.
Figure 2/E.412 - T0200990-87
3.2.2 Skip
This control allows traffic to bypass a specified circuit group and
advance instead to the next circuit group in its normal routing pattern.
Typical application: Used to bypass a congested circuit group or distant
exchange when the next circuit group can deliver the call attempts to the
destination without involving the congested circuit group or exchange.
Application is usually limited to networks with extensive alternative
routing. When used on both-way circuit groups it has an expansive effect
on traffic flow in the opposite direction.
3.2.3 Temporary alternative routing
This control redirects traffic from congested circuit groups to other
circuit groups not normally available which have idle capacity at the time.
Typical application: To increase the number of successful calls during
periods of circuit group congestion and to improve the grade of service to
subscribers.
3.2.4 Special recorded announcements
These are recorded announcements which give special information to
operators and/or subscribers, such as to defer their call to a later time.
Typical application: Used to notify customers of unusual network
conditions, and to modify the calling behavior of customers and operators
when unusual network conditions are present. Calls that are blocked by
other network management controls can also be routed to a recorded
announcement.
Figure 3/E.412 - T0201010-87
4 Automatic exchange controls
Automatic dynamic network management controls represent a significant improvement over
conventional controls. These controls, which are preassigned, can
quickly respond to conditions internally detected by the exchange,
or to status signals from other exchanges, and are promptly removed
when no longer required. Automatic control applications should be
planned, taking into account the internal overload control strategy
provided in the exchange software.
4.1 Automatic congestion control system
4.1.1 Exchange congestion
When a digital international/transit exchange carries traffic above the
engineered level, it can experience an overload that diminishes its total call
processing capability. Because of the speed of the onset of such congestion and
the critical nature of the condition, it is appropriate that control be
automatic. The automatic congestion control (ACC) system consists in the
congested exchange sending a congestion indicator to the connected exchange(s)
using common channel signalling. The exchange(s) receiving the congestion
indication can respond by reducing a certain percentage of the traffic offered to
the congested exchange, based on the response action selected for each
application.
Fascicle II.3 - Rec. E.412 PAGE1
4.1.2 Detection and transmission of congestion status
An exchange should establish a critical operating system benchmark, and
when continued levels of nominal performance are not achieved (e.g. due to
excessive traffic), a state of congestion is declared. Thresholds should be
established so that the two levels of congestion can be identified, with
congestion level 2 (CL2) indicating a more severe performance degradation than
congestion level 1 (CL1). When either level of congestion occurs, the exchange
should have the capability to:
1) code an ACC indication in the appropriate common channel signalling
messages, and
2) notify its network management centre and support system of a change in
its current congestion status.
4.1.3 Reception and control
When an exchange receives a signal that indicates a congestion problem at
a connected exchange, the receiving exchange should have the capability to reduce
the number of seizures sent to the congested exchange.
An exchange should have the capability of:
1) assigning an ACC response action on an individual circuit group1)
basis, as specified by the network manager, and
2) notifying its network management centre and support system of a change
in congestion status received from a distant exchange.
There should be several control categories available in the exchange. Each
category would specify the type and amount of traffic to be controlled in
response to each of the received ACC indicators. The categories could be
structured so as to present a wide range of response options.
For a specific ACC response category, if the received ACC indicator is set
to a CL1 condition then the receiving exchange could, for example, control a
percentage of the Alternate Routed To (ART) traffic to the affected exchange. The
action taken by the control would be to either SKIP or CANCEL the controlled
calls, depending on the ACC response action that was assigned to that circuit
group. In a similar manner, if a CL2 condition is indicated, then the receiving
exchange could control all ART traffic and some percentage of Direct Routed (DR)
traffic. Other options could include the ability to control hard-to-reach
traffic, or transit traffic. In the future, control categories could be expanded
to include service-specific controls. This would be particularly useful in the
transition to ISDN.
Note - ACC response categories can be set locally in the exchange or by
input from a network management centre, or operations system.
Table 1/E.412 is an example of the flexibility that could be achieved in
response to a signal from an exchange that is experiencing congestion. In this
example, different control actions would be taken based upon the distinction
between ART and DR traffic types. These actions could represent the initial
capabilities available with the ACC control. Other alternatives in the future
could include the ability to control hard-to-reach traffic (see S 2.2), or
transit traffic or to provide other controls such as call-gapping. Additional
response categories could also be added to Table 1/E.412 to give greater
flexibility and more response options to the ACC control. It could also be
possible to exclude priority calls from ACC control.
TABLE 1/E.412
ACC control response
Congestion level Traffic type Response category
A B C
CL1 ART 0 0 100
DR 0 0 0
CL2 ART 100 100 100
1) In this context, the term "circuit group" refers to all of the outgoing and
both-way circuit sub-groups which may directly connect the congested exchange and the
responding exchange.
PAGE2 Fascicle II.3 - Rec. E.412
DR 0 75 75
4.1.4 Any international application of ACC should be based on negotiation and
bilateral agreement among the affected Administrations. This includes an
agreement as to whether the controlled calls should be skipped or cancelled.
Application within a national network would be a national matter. An exchange
that is capable of "ACC receive and control" should not indiscriminately assign
ACC to all routes since a distant exchange may be equipped for common channel
signalling, but may not yet have an ACC transmit capability. This could result in
invalid information in the ACC fields in the signalling messages and the
inappropriate application of ACC controls at the receiving exchange. Additional
details on the ACC system are in Recommendation Q.542.
4.2 Selective circuit reservation control
4.2.1 The selective circuit reservation control enables an exchange to
automatically give preference to a specific type (or types) of traffic over
others (e.g., direct routed calls over alternate routed calls) at the moment when
circuit congestion is present or imminent. The selective circuit reservation
control can be provided with one or two thresholds, with the latter being
preferred due to its greater selectivity. Specific details on the selective
circuit reservation control may be found in Recommendation Q.542.
4.2.2 General characteristics
The selective circuit reservation control has the following operating
parameters:
- a reservation threshold(s),
- a control response,
- a control action option.
The reservation threshold defines how many circuits or how much circuit
capacity should be reserved for those traffic types to be given preferred access
to the circuit group. The control response defines which traffic types should be
given a lesser preference in accessing the circuit group, and the quantity of
each type of traffic to control. The control action option defines how those
calls denied access to the circuit group should be handled. The control action
options for processing of calls denied access to the circuit group may be SKIP or
CANCEL.
When the number of idle circuits or the idle capacity in the given circuit
group is less than or equal to the reservation threshold, the exchange would
check the specified control response to determine if calls should be controlled.
The SKIP response allows a call to alternate-route to the next circuit group in
the routing pattern (if any) while the CANCEL response blocks the call.
These parameters should be able to be set locally in the exchange for each
selected circuit group or by input from a network management operations system.
In addition, the network manager should have the capability to enable and disable
the control, and to enable the control but place it in a state where the control
does not activate (e.g., by setting the reservation threshold to zero). Further,
the network manager should have the ability to set the values for the response
categories.
4.2.3 Single threshold selective circuit reservation control
In this version of the control, only a single reservation threshold would
be available for the specified circuit group.
Table 2/E.412 is an example of the flexibility that could be achieved in
the control's response to circuit group congestion. In the future, other
distinctions between traffic could be identified that would expand the number of
traffic types in Table 2/E.412. An example would be to control hard-to-reach
traffic as indicated in S 2.2, or to give preference to priority calls.
4.2.4 Multi-threshold selective circuit reservation control
The multi-threshold control provides two reservation thresholds for the
specified circuit group. The purpose of multiple reservation thresholds is to
allow a gradual increase in the severity of the control response as the number of
idle circuits in the circuit group decreases. The only restriction on the
assignment of reservation thresholds would be that a reservation threshold
associated with a more stringent control must always be less than or equal to the
reservation threshold of any less stringent control, in terms of the number of
reserved circuits, or circuit capacity.
TABLE 2/E.412
An example of a single threshold selective circuit reservation Percentage control response
Fascicle II.3 - Rec. E.412 PAGE1
table
Circuit group Traffic type Response category assigned to
reservation circuit group
threshold A B C
RT1 ART 25 50 100
DR 0 0 25
Table 3/E.412 is an example of the flexibility that could be achieved in
the control's response to circuit group congestion with a two-reservation
threshold control. In the future, other distinctions between traffic could be
identified that would expand the number of traffic types in Table 3/E.412. An
example would be to control hard-to-reach traffic as indicated in S 2.2.
TABLE 3/E.412
An example of a two-threshold selective circuit reservation Percentage control response
table
Circuit group Traffic type Response category assigned to circuit
reservation group
threshold A B C D E
RT1 ART
PAGE2 Fascicle II.3 - Rec. E.412
25 50 75 100 100
DR 0 0 0 0 0
RT2 ART 50 75 75 100 100
DR 0 0 25 50 100
Fascicle II.3 - Rec. E.412 PAGE1
5 Status and availability of network management controls
5.1 The exchange and/or network management operations systems should provide
information to the network management centre and/or the exchange staff as to what
controls are currently active and whether the controls were activated
automatically or by human intervenion. Measurements of calls affected by each
control should also be available (see Recommendation E.502).
5.2 To help insure the viability of network management functions during
periods of exchange congestion, network management terminals (or exchange
interfaces with network management operations systems), and functions such as
controls, should be afforded a high priority in the exchange operating software.
6 Operator controls
Traffic operators are usually aware of problems as they occur in the
network, and this information can reveal the need to control traffic. The
operators can then be directed to modify their normal procedures to reduce
repeated attempts (in general, or only to specified destinations), or to use
alternative routings to a destination. They can also provide information to
customers and distant operators during unusual situations, and can be provided
with special call handling procedures for emergency calls.
PAGE2 Fascicle II.3 - Rec. E.412
