Sourced in Canada.
The information in this document is subject to change without notice. The statements, configurations, technical
data, and recommendations in this document are believed to be accurate and reliable, but are presented without
express or implied warranty. Users must take full responsibility for their applications of any products specified in this
document. The information in this document is proprietary to Nortel Networks.
Nortel, the Nortel Logo, the Globemark, SL-1, Meridian 1, and Succession are trademarks of Nortel Networks.
All other trademarks are the property of their respective owners.
Page 3
Revision History
May 2007
Standard 01.01. This document is issued to support Communication Server
1000 Release 5.0. This document contains information previously contained
in the following legacy document, now retired: (553-3001-363).
August 2005
Standard 3.00. This document is up-issued for Communication Server 1000
Release 4.5.
September 2004
Standard 2.00. This document is up-issued for Communication Server 1000
Release 4.0.
October 2003
Standard 1.00. This document is a new NTP for Succession 3.0. It was
created to support a restructuring of the Documentation Library. This
document contains information previously contained in the following legacy
document, now retired: IP Trunk: Description, Installation, and Operation
(553-3001-202).
Getting help from the Nortel web site17
Getting help over the telephone from a Nortel Solutions Center17
Getting help from a specialist by using an Express Routing Code17
Getting help through a Nortel distributor or re-seller18
Overview of IP Trunk 3.0119
Contents 19
Introduction 19
Startup and registration 23
IP Trunk 3.01 (and later) and CS 1000M25
IP Trunk 3.01 (and later) requirements27
Interoperability with the ITG 8-port trunk card 28
5
Loss plans and pad values 27
Codec selection27
Package requirements27
TM 3.128
System description29
Contents 29
IP Trunk 3.01 (and later) application31
System requirements32
Hardware components for IP Trunk 3.01 (and later)34
Ordering rules and guidelines36
Ordering rules for an IP Trunk 3.01 (and later) node36
Ordering rules for IP Trunk 3.01 (and later) node expansion37
Sparing ratios for IP Trunk 3.01 (and later) components37
Size of the IP Trunk 3.01 (and later) network104Endpointtype105TheAverage Hold Time (AHT) and distribution of incoming calls105CalculateEthernet and WAN bandwidth usage112SilenceSuppression engineering considerations114Faxengineering considerations115TAT andTROconsiderations 116
WANroute bandwidth engineering119
Assess WAN link resources122
Link utilization122Estimate network loading caused by IP Trunk 3.01 (and later) traffic 123Route Link Traffic Estimation124Enough capacity126Insufficient link capacity 127Other intranet resource considerations127
Implement QoS in IP networks127
Traffic mix 128TCP traffic behavior 128IP Trunk 3.01 (and later) DiffServ support for IP QoS 129Queue management130
Network delay and packet loss evaluation example 147
Other measurement considerations148
Estimate voice quality148
Does the intranet meet expected IP Trunk 3.01 (and later) QoS?153
IP Trunk 3.01 (and later) LAN installation and configuration154
Basic setup of the IP Trunk 3.01 (and later) system154
IP trunk card connections154
Configure a system with separate subnets for voice and management155
Subnet configurations155
Selecting public or private IP addresses 157
Single subnet option for voice and management157
Multiple IP Trunk 3.01 (and later) nodes on the same ELAN and
General LAN considerations158ELAN and TLAN network interface half- or full-duplex operation158TLAN subnet design 159Configure the TLAN subnet IP router 159Setting up the ELAN subnet160How to avoid system interruption160
Install filter and NTND26 cable (for MSDL and DCHIP cards in different Large
System equipment rows) 202
Small System cable installation 203
Install the serial cable 204
Cabling for the Media Card 32-port trunk card 205
ELAN and TLAN network interfaces 205
ITG Card ELAN/TLAN Adapter (L-adapter) 206
RS-232 maintenance port210
NTMF29BA DCHIP cable 211
DCHIP cable routing, Large Systems 212
DCHIP Cable Routing
Meridian 1 Option 11C Cabinet/CS 1000M Cabinet213
Other components214
Media Card 32-port trunk card modem connection215
Configure IP Trunk 3.01 (and later) data 216
Configure the ISL D-channel on the system for the DCHIP card for IP Trunk
3.01 (and later) 216
Configure the ISL D-channel on the Meridian 1/CS 1000M for the DCHIP card
for IP Trunk 3.01 (and later)219
Configure ISDN feature in Customer Data Block220
Configure IP Trunk 3.01 (and later) TIE trunk routes221
Configure Media Card 32-port and ITG-Pentium 24-port trunk cards and units for
IP Trunk Route225
Configure dialing plans within the corporate network228
Make the IP Trunk 3.01 (and later) the first-choice, least-cost entry in the Route
List Block228
Turn on Step Back on Congestion for the IP Trunk 3.0 (and later) trunk route229
Turn off IP Trunk 3.01 (and later) route during peak traffic periods on the IP data
network 229
ESN5 network signaling 229
Disable the Media Card 32-port and ITG-Pentium 24-port trunk cards 234
Configure IP Trunk 3.01 (and later) data in TM 3.1234
Add an IP Trunk 3.01 (and later) node in TM 3.1 manually 235
Add an IP Trunk 3.01 (and later) node and configure general node
properties235
Single vs. separate TLAN and ELAN subnets 237
Configure Network Connections237
Configure card properties 239
Configure DSP profiles for the IP Trunk 3.01 (and later) node242
Configure SNMP Traps/Routing and IP addresses tab 246
Configure Accounting server 249
Control node access with SNMP community name strings250
Exit node property configuration session251
Create the IP Trunk 3.01 (and later) node dialing plan using TM 3.1251
Retrieve the IP Trunk 3.01 (and later) node dialing plan using TM 3.1257
Transmit IP trunk card configuration data from TM 3.1 to the IP trunk cards259
Before configuration data is transmitted 259
Configure the Leader 0 IP address259
Backup Leader installation for IP Trunk 3.01 (and later)261
Transmit the node properties, card properties and dialing plan to Leader 0263
Verify installation and configuration 265
Observe IP Trunk 3.01 (and later) status in TM 3.1 265
Transmit card properties and dialing plan to Leader 1 and Follower cards267
Configure date and time for the IP Trunk 3.01 (and later) node268
Change the default ITG shell password to maintain access security269
Change default ESN5 prefix for non-ESN5 IP telephony gateways 270
Check and download IP trunk card software in TM 3.1 271
Transmit new software to the IP trunk cards 273
Upgrade the DCHIP PC Card275
Configure TM 3.1 Alarm Management to receive SNMP traps from the IP trunk
cards 276
Make test calls to the remote nodes (ITG Trunk or IP Trunk) 279
Provisioning IP Trunk 3.01 (and later) in TM 3.1281
Contents 281
Overview 281
Add a site and system282
Add a site 282
Change an existing site284
Delete a site 286
Add a system 289
Delete a system 299
Add an IP Trunk 3.01 (and later) node301
Edit a node 311
Delete a node 316
Define the dialing plan information 318
Non-Gatekeeper-resolved (local) dialing plan 318
Gatekeeper-resolved endpoints333
TM 3.1 OA and M using TM 3.1 applications341
Contents 341
Introduction 342
TM 3.1 OA and M procedure summary342
Delete a node 343
Delete an IP trunk card343
Database locking344
ITG Card Properties window345
ITG Card Properties Maintenance window345
ITG Card Properties Configuration window347
Add an IP Trunk 3.01 (and later) node on TM 3.1 by retrieving an existing node351
Retrieve and add an IP Trunk 3.01 (and later) node for administration
purposes351
Retrieve and add an IP Trunk 3.01 (and later) node for maintenance and diagnostic
purposes353
Configuration audit354
Retrieve IP Trunk 3.01 (and later) configuration information from the IP Trunk
3.0 (and later) node 355
Schedule and generate and view IP Trunk 3.01 (and later) OM reports356
System commands LD 32360
Disable the indicated IP trunk card361
Disable the indicated IP trunk card when idle362
Enable an indicated IP trunk card362
Disable an indicated IP trunk card port362
Enable an indicated IP trunk card port362
Display IP trunk card ID information 362
Display IP trunk card status362
Display IP trunk card port status363
OA and M using the ITG shell CLI and overlays365
Contents 365
Introduction 366
ITG Shell OA and M procedure summary 366
Access the ITG shell through a maintenance port or Telnet366
Connect a PC to the card maintenance port367
Telnet to an IP trunk card through the TM 3.1 PC368
Change the defaultITG shell password to maintain access security369
Reset the default ITG shell password370
Download the ITG operational measurements through the ITG shell372
Reset the operational measurements372
Display the number of DSPs 373
Display IP Trunk 3.01 (and later) node Properties373
Display IP Trunk 3.01 (and later) Gatekeeper status374
Transfer files through the Command Line Interface375
Upgrade IP trunk card software using FTP 377
Backup and restore from the CLI 380
Recover the SNMP community names 381
IP Trunk 3.01 (and later) configuration commands 382
Download the IP Trunk 3.01 (and later) error log382
Disable the indicated IP trunk card when idle384
Enable an indicated IP trunk card384
Disable an indicated IP trunk card port384
Enable an indicated IP trunk card port385
Display IP trunk card ID information 385
Display IP trunk card status385
Display IP trunk card port status385
Maintenance387
Contents 387
Introduction 388
IP Trunk 3.01 (and later) IP trunk card alarms 389
System level maintenance394
Access the IP trunk card394
IP trunk card LD commands395
TM 3.1 maintenance commands 396
Multi-purpose Serial Data Link (MSDL) commands 397
Simple Network Management Protocol (SNMP) 397
TRACE and ALARM/LOG398
ITG shell command set 398
IP trunk card self-tests406
This chapter explains how to get help for Nortel products and services.
Getting help from the Nortel web site
The best way to get technical support for Nortel products is from the Nortel
Technical Support web site:
ttp://www.nortel.com/support
h
This site provides quick access to software, documentation, bulletins, and
tools to address issues with Nortel products. From this site, you can:
•
download software, documentation, and product bulletins
•
search the Technical Support Web site and the Nortel Knowledge Base
for answers to technical issues
•sign up for automatic notification of new software and documentation
for Nortel equipment
•
open and manage technical support cases
17
Getting help over the telephone from a Nortel Solutions Center
If you do not find the information you require on the Nortel Technical Support
web site, and you have a Nortel support contract, you can also get help over
the telephone from a Nortel Solutions Center.
In North America, call 1-800-4NORTEL (1-800-466-7835).
Outside North America, go to the following web site to obtain the telephone
number for your region:
h
ttp://www.nortel.com/callus
Getting help from a specialist by using an Express Routing Code
To access some Nortel Technical Solutions Centers, you can use an Express
Routing Code (ERC) to quickly route your call to a specialist in your Nortel
product or service. To locate the ERC for your product or service, go to:
Getting help through a Nortel distributor or re-seller
If you purchased a service contract for your Nortel product from a distributor
or authorized re-seller, contact the technical support staff for that distributor
or re-seller.
"Interoperability with the ITG 8-port trunk card" (page 28)
Introduction
The IP Trunk 3.01 (and later) software application is an Internet Telephony
Gateway (ITG) trunk software application that maintains the functionality of
ITG Trunk 2.x using Integrated Services Digital Network (ISDN).
19
IP Trunk 3.01 (and later) allows networks with Meridian 1 IP-enabled
systems to add a CS 1000 system to the existing IP Telephony network.
This increases the range of system options to provide enterprise-wide
telephony services.
IP Trunk 3.01 (and later) provides call-routing flexibility and survivability.
Even with a Signaling Server acting as a centralized authority for routing
IP Telephone calls, IP Trunk can make some call-routing decisions locally.
This can be done for one of the following reasons:
•
It can maintain at least a minimum level of service in the unlikely event
that all Signaling Servers on the network are unreachable.
•
It can maintain the existing functionality within a pre-existing ITG Trunk
network that was upgraded to IP Trunk 3.01 (and later).
In addition to routing IP Telephony calls with locally configured call-routing
options, IP Trunk 3.01 takes advantage of the centralized IP Telephony call
routing of an H.323 Gatekeeper residing on a Signaling Server elsewhere
on the network.
The H.323 Gatekeeper allows or denies access to IP network gateways. It
also provides address analysis to find the destination gateway or device.
A gateway is a device that translates circuit-switched signaling into H.323
signaling and translates circuit-switched bit stream user data into packetized
user data to enable the data to be delivered across an IP network. IP Trunk
3.01 (and later) provides IP access between the Meridian 1/CS 1000M
system and the IP network carrying voice traffic.
IP Trunk 3.01 (and later) interworks with ITG Trunk 2.x, but not with ITG
Trunk 1.0. For ITG Trunk 1.0 to interwork with IP Trunk 3.01 (and later),
upgrade ITG Trunk 1.0 to ITG Trunk 2.0. See Appendix "Upgrade an ITG
Trunk 1.0 node to support ISDN signaling trunks" (page 471).
IP Trunk 3.01 (and later) interworks with a CS 1000M system, which fulfils
the role of a Gatekeeper. The Gatekeeper uses directly-routed calls. See
"Directly-routed calls" (page 22). Using H.323 Registration and Admission
Signaling (RAS), IP Trunk 3.01 (and later) registers with the Gatekeeper,
if provisioned to do so. IP Trunk 3.01 (and later) then processes calls by
scanning its directory number information and routes unresolved calls to
the Gatekeeper.
For a Meridian 1 system to interwork with a CS 1000M system, the following
requirements must be met:
•
The ITG-Pentium 24-port trunk card and the Media Card 32-port trunk
card must be upgraded to IP Trunk 3.01 (and later) software. This
upgrade supports MCDN features and Gatekeeper registration. As well
as this document, see Telephony Manager 3.1 System Administration(NN43050-601) for more information on installing, upgrading, and
upgrading IP Trunk 3.01 (and later) parameters.
•
The IP Trunk 3.01 (and later) node must be configured to register with
the CS 1000M Gatekeeper. Refer to "Gatekeeper-resolved endpoints"
(page 333) and to Telephony Manager 3.1 System Administration
(NN43050-601) for more information on how to configure the IP Trunk
3.01 (and later) options.
IP Trunk 3.01 (and later) is subordinate to the Gatekeeper for all calls that
require Gatekeeper intervention. This means that the IP Trunk 3.01 (and
later) node performs the following actions:
handles the call based on the return message from the Gatekeeper
IP Trunk 3.01 (and later) accesses additional devices through the
Gatekeeper. It is no longer necessary to individually provision the entire
mesh at each IP Trunk 3.01 (and later) node. Instead, the calls go to the
Gatekeeper, which provides the IP Trunk 3.01 (and later) application with
the correct destination for the call. See Figure 1 "IP Trunk 3.01 (and later)
architecture" (page 21).
Figure 1
IP Trunk 3.01 (and later) architecture
Introduction 21
IP Trunk 3.01 (and later) uses the Meridian 1/CS 1000M core switch as the
primary driver, which sends ISDN messages through the ISDN Signaling
Link (ISL) to the IP trunk card for IP Trunk 3.01 (and later) processing. IP
Trunk 3.01 (and later) tandems the Meridian 1/CS 1000M core switch to the
IP network, providing point-to-multipoint connection.
Alternatively, depending on the provisioning and the requested destination,
if a call cannot be resolved locally, IP Trunk 3.01 (and later) can interwork
with the Gatekeeper to identify the destination node before routing directly
to that destination.
Two types of calls can be routed through interworking with the Gatekeeper:
directly-routed calls and Gatekeeper-routed calls.
In directly-routed calls, the Gatekeeper returns the IP address of the call’s
actual destination.
Figure 2 "Directly-routed call" (page 22) on Figure 2 "Directly-routed call"
(page 22) represents a directly-routed call. Once the destination IP address
is obtained, the originator sends the call directly to the destination node.
Figure 2
Directly-routed call
WARNING
The only Gatekeeper that IP Trunk 3.01 (and later) officially
supports is the CS 1000M Gatekeeper. Gatekeeper calls made
between the CS 1000M system and IP Trunk 3.01 (and later) are
directly-routed calls.
Gatekeeper-routed calls
In Gatekeeper-routed calls, the Gatekeeper returns the Gatekeeper’s IP
address and port as both the destination for the originating call and the
originator for the destination, rather than the end-point address and port.
Figure 3 "Gatekeeper-routed call" (page 23) represents a Gatekeeper-routed
call. The destination IP address provided by the Gatekeeper is the
Gatekeeper’s IP address. All messages are routed through the Gatekeeper.
On system startup, the IP Trunk 3.01 (and later) Leader card is established,
based on whether the primary and backup Leaders come up, in what
sequence, and how quickly. This operation remains unchanged from prior
releases. It provides all necessary information to the follower cards.
Part of the information in the Dial Plan table is the Gatekeeper registration
information, which includes three main fields: the local node H.323 identifier
(node name), a flag indicating registration handling, and a third field for
future development.
The registration handling has two potential flag values as follows:
•
0 – Register the IP addresses of all cards (Leader 0, Leader 1, and
Follower cards) in the IP Trunk 3.01 (and later) node.
•
1 – Each card must register individually, if required. When registering
with a CS 1000M Gatekeeper, IP Trunk 3.01 (and later) registers only
the node address. No other IP addresses are sent to the Gatekeeper
in the Registration Request (RRQ) message.
The flag value is ignored when the provisioned Gatekeeper is a CS
1000M Gatekeeper.
On startup, if the IP Trunk 3.01 (and later) Leader is provisioned to use a
Gatekeeper, it seeks out and locates the Gatekeeper using RAS signalling
and then registers with the Gatekeeper using an RRQ. As part of the
registration process, the IP Trunk 3.01 (and later) Leader registers using the
registration handling flag to determine how to proceed.
The Gatekeeper and IP Trunk 3.01 (and later) re-register on a regular basis,
based on the Time To Live (TTL) configured for the IP path.
The Gatekeeper is the final authority on the TTL values. The Gatekeeper
can override the provisioned value of IP Trunk 3.01 (and later) and require
the IP Trunk 3.01 (and later) gateway to change its TTL value to match that
required by the Gatekeeper.
Depending on the Gatekeeper type (for example, Gatekeepers other than
CS 1000M), if the Gatekeeper flag in the dial plan file indicates the need for
multiple IP Trunk 3.01 card IP addresses (flag value = 0), the
all IP addresses for the node. These additional IP addresses are reserved
exclusively for calls to the Gatekeeper. By sending all the IP addresses in
the RRQ, the Gatekeeper is able to determine the origin of the admission
requests. These addresses are used when the Gatekeeper considers
the endpointIdentifier sent to the gateway in the RRQ confirmation to
be insufficient to confirm that the Admission Request (ARQ) belongs to a
gateway registered with that Gatekeeper. The Gatekeeper rejects any ARQ
from an unknown end-point.
RRQ includes
CS 1000M requires an endpointIdentifier match and does not care about
the IP addresses. Therefore, the Gatekeeper flag is unnecessary for CS
1000M.
On startup, the message flow between the IP trunk card serving as the IP
Trunk 3.01 (and later) Active Leader and the Gatekeeper is as follows:
1. Gatekeeper Request (GRQ) – From the Active Leader to the
Gatekeeper, using the provisioned Gatekeeper IP address. The Optivity
Telephony Manager (TM 3.1) configuration indicates where the IP
Trunk 3.01 (and later) node must look for its Gatekeeper, but this is
not necessarily the actual Gatekeeper address the node uses for call
processing.
Some Gatekeepers use a "virtual IP address" to screen the fact that
the Gatekeeper with which the gateway registers has internal standby
controllers. In this case, the request might go to a Gatekeeper server
that determines the correct virtual IP address. The Gatekeeper’s
internal Message Forwarding process sends the messages to the
current active Gatekeeper node.
CS 1000M do not require a Gatekeeper Request from IP Trunk 3.01
(and later); therefore, no Request or Confirm is sent.
2. Gatekeeper Confirm (GCF) – From the Gatekeeper to the Active
Leader,with the functional Gatekeeper IP address. This address is used
for all call control messaging and registration messages between the IP
Trunk 3.01 (and later) cards and the Gatekeeper.
3. Gatekeeper Registration Request (RRQ) – From the Active Leader to
the Gatekeeper, with all of the node’s IP addresses.
IP addresses are only sent if required. A CS 1000M does not require all
IP addresses, so the IP addresses are not sent.
4. Gatekeeper Register Confirm (RCF) – From the Gatekeeper to the
Active Leader, providing the TTL prior to a re-registration attempt by
the leader and indicating under what conditions admission requests
are needed.
Typically, the TTL is in minutes. The default IP Trunk 3.01 (and later)
value, if no response from the Gatekeeper is received, is 300 seconds.
However, the Gatekeeper can enforce a shorter interval in seconds or
tens of seconds. The standards allow seconds from 1 to (232) –1.
Recommendation
Nortel recommends that the TTL be provisioned in the 30- to 60-second
range.
The IP Trunk 3.01 (and later) node must perform a "keep-alive"
re-registration prior to the expiry of the timer on the Gatekeeper. When
the Gatekeeper timer expires, a full registration is needed.
IP Trunk 3.01 (and later) and CS 1000M
The CS 1000M systems use virtual trunking (IP Peer Networking) to
inter-operate with the IP Trunk 3.01 (and later) nodes. However, the CS
1000M can be a Gatekeeper for the system.
When IP Trunk 3.01 (and later) is part of a network with a Signaling Server
acting as a central control point, it is able to take partial advantage of a
feature known as IP Peer Networking. IP Peer Networking eliminates the
multiple conversions between IP and non-IP circuits, increasing call routing
efficiency and overall voice quality. Many calls involving an IP Peer endpoint
and one or more IP Trunk endpoints can use this capability. However, calls
that use only IP Trunk facilities, and a small subset of calls involving both
IP Trunk and IP Peer, cannot obtain this benefit.
IP Trunk 3.01 (and later) supports Gatekeeper Registration and Admission
Signaling (RAS) and Call Admission Signaling. IP Trunk 3.01 (and later)
interworks with CS 1000M, which fulfills the role of a Gatekeeper. Using
H.323 RAS, IP Trunk 3.01 (and later) uses RAS Messaging to register
with the Gatekeeper if provisioned to do so. IP Trunk 3.01 (and later) then
processes calls by scanning its Directory Number (DN) information. If the
call is not resolved using the local Address Translation Protocol Module
(ATPM) and IP Trunk 3.01 (and later) is registered with a Gatekeeper, then
IP Trunk 3.01 (and later) routes the call to the Gatekeeper.
The IP Trunk 3.01 (and later) node is subordinate to the Gatekeeper for all
calls requiring the Gatekeeper. The IP Trunk 3.01 (and later) node registers
with the Gatekeeper according to H.323 protocol, requests admission,
accepts the reply according to H.323 protocol, and handles the call based
on the returned message from the Gatekeeper.
A CS 1000M node consists of two components:
•
Call Server – used for call control of CS 1000M gateways
•
Signaling Server – used for protocol analysis
The CS 1000M Gatekeeper accepts the registration of multiple IP trunk
cards implicitly in a single RRQ. This means that all Follower cards are
registered at the same time as the Leader card, because the CS 1000M
node returns an endpointIdentifier assigned by the Gatekeeper to that
node. Later, a request to establish a call to a Gatekeeper-controlled
endpoint receives in the response the enpointIdentifier of the endpoints
that was provided at registration.
The CS 1000M gateways interwork with the IP Trunk 3.01 (and later)
gateway resident function which generates the FACILITY redirect. The
FACILITY redirect is used when calls terminate at an IP Trunk 3.01 (and
later) node. The CS 1000M gateways do not use this redirection themselves.
Other Gatekeepers accept the FACILITY redirect and registration of multiple
IP trunk cards in a single RRQ; that is, the Followers are registered with,
and at the same time as, the Leader.
IP Trunk 3.01 (and later) interworks with the CS 1000M systems and IP
Peer Networking. As CS 1000M and IP Peer Networking use MCDN only,
the only applicable protocol is MCDN. IP Trunk 3.01 (and later) uses the
"interoperability format" of the non-standard data with IP Peer Networking
and all other gateways accessible through CS 1000M.
When IP Trunk 3.01 (and later) inter-operates with itself, with ITG Trunk
2.x.25, or with BCM 2.5 FP1, the IP Peer Networking CS 1000M Gatekeeper
is not required. The existing ITG Trunk 2.1 node-based dialing plan is
converted automatically to IP Trunk 3.01 (and later) by .
There are no direct media paths between the Meridian 1 telephones and the
CS 1000M telephones. There are direct paths between the IP Trunk 3.01
(and later) IP trunk cards and the CS 1000M telephones.
When the IP Trunk card is in a CS 1000 system, it can take advantage
of the Dynamic Loss Plan developed for the IP Peer product. This allows
the system core to inform the IP Trunk card of the correct pad levels to be
used. As with IP Peer, it also allows the creation of a custom table when the
environment requires one.
When using Dynamic Loss Plan, the node must be provisioned to have a
default loss plan pad of 0 in both the transmit and receive directions. This
allows a 0 transmit and receive level when the IP Trunk has a tandem to
another trunk device, improving voice quality.
Codec selection
A CS 1000M network is generally designed for use with a G.711 Codec.
In cases where minimizing bandwidth usage in a CS 1000M network is a
consideration, G.729 might be used.
Nortel recommends provisioning G.711 Codec in IP Trunk 3.01 (and later) and
in all other network equipment to facilitate communication with CS 1000M.
IP Trunk 3.01 (and later) requirements27
Recommendation
IP Trunk 3.01 (and later) requirements
IP Trunk 3.01 requires a minimum of Release 25.15 software. To interwork
with the CS 1000M Gatekeeper, CS 1000 Release 3.0 software (or later)
is required.
package requirements for the IP Trunk 3.01 (and later) application.
Unlike ITG Trunk 2.0, Q-Signaling protocol (QSIG) support is not required in
IP Trunk 3.01 (and later), though it is available for Large Systems. Meridian
1 Option 11C Cabinet, CS 1000M Cabinet, Meridian 1 PBX 11C Chassis,
and CS 1000M Chassis do not support QSIG signaling. Therefore, the
Multi-purpose Serial Data Link (MSDL), applicable only to Large Systems,
is recommended but not mandatory; the earlier D-channel interface cards
can provide Meridian Customer Defined Network (MCDN) ISDN Signaling
Link (ISL). QSIG and MSDL are incompatible for feature transport. If both
QSIG and MSDL are configured on the network, this can cause the loss of
features such as Name Display, Ring Again, and Transfer Notification and
subsequent path simplification operations.
Table 1
IP Trunk 3.01 (and later) package requirements
Package
Name
BARS
NARS
CDP
ISDN
ISL
NTWK
FNP
Package
Number
57
58
59
145
147
148
160
Package description
Basic Alternate Route
Selection
Network Alternate Route
Selection
Coordinated Dialing PlanRequired if Dialing Plan used.
ISDN BaseMandatory. No D-channel can exist
ISDN Signaling LinkMandatory. ISL cannot exist without
Advanced ISDN Network
Services
Flexible Numbering PlanRequired if Dialing Plan used.
Comments
Package 57 and/or 58 is required.
Package 57 and/or 58 is required.
If the configuration restricts NARS,
use CDP to obtain private network
dialing. CDP can also co-exist with
NARS.
without this package.
this package. Without ISL, the
Meridian 1/CS 1000M to IP Trunk
D-channel cannot be provisioned.
Required if Networking Services
used.
When the configuration allows CDP,
FNP is recommended, but not
mandatory.
MSDL
222
Multipurpose Serial Data
Link
Recommended for MSDL on Large
systems.
TM 3.1
TM 3.1 is required to configure and maintain IP Trunk 3.01 (and later).
Interoperability with the ITG 8-port trunk card
Telephone calls can be made between IP Trunk 3.01 (and later) and ITG
Trunk 2.x.
This section contains information on the following topics:
"IP Trunk 3.01 (and later) application" (page 31)
"System requirements" (page 32)
"Hardware components for IP Trunk 3.01 (and later)" (page 34)
"Ordering rules and guidelines" (page 36)
"Ordering rules for an IP Trunk 3.01 (and later) node" (page 36)
"Ordering rules for IP Trunk 3.01 (and later) node expansion" (page 37)
"Sparing ratios for IP Trunk 3.01 (and later) components" (page 37)
"E-Model" (page 73)
"Fallback to alternate facilities" (page 74)
"Triggering fallback to alternate trunk facilities" (page 74)
"Fallback in IP Trunk 3.01 (and later)" (page 75)
"Return to the IP network" (page 76)
"Type of Service" (page 76)
"Fax support" (page 78)
"Remote Access" (page 79)
"Per-call statistics support using RADIUS Client" (page 80)
IP Trunk 3.01 (and later) supports ISDN Signaling Link (ISL) IP trunks on
the Media Card 32-port trunk card and the ITG-Pentium 24-port trunk card.
The NTCW80 8-port trunk card cannot be upgraded to IP Trunk 3.01 (and
later).
An ISDN Signaling Link D-Channel (ISL DCH) provides DCH connectivity
to the system and signaling control for the ports on the IP trunk card and
any additional ports on other IP trunk cards in the same node. The DCH
connection expands the signaling path between the Meridian 1/CS 1000M
and the gateway. IP Trunk 3.01 (and later) allows Meridian 1/CS 1000M
systems to be networked using ISDN, while transmitting H.323 signaling
and voice over a standard IP protocol stack.
IP Trunk 3.01 (and later) application31
IP Trunk 3.01 (and later) compresses voice and demodulates Group 3 Fax.
IP Trunk 3.01 (and later) then routes the packetized data over a private
IP network.
IP Trunk 3.01 (and later) delivers an ISDN signaling interface between
the Meridian 1 and the Voice (and fax) over IP (VoIP) interface. The high
signaling bandwidth of this ISDN interface expands the feature functionality
for VoIP trunks. It provides, for example, Calling Line Identification (CLID)
and Call Party Name Display (CPND).
To install IP Trunk 3.01 (and later), the customer must have a corporate IP
network with managed bandwidth capacity, and routers available for WAN
connectivity between networked Meridian 1/CS 1000 systems. The best
VoIP performance is obtained with a QoS-managed network.
The LAN connection of IP Trunk 3.01 (and later) requires 10BaseT or
100BaseTX Ethernet network interfaces for voice (TLAN network interface)
and 10BaseT for management and D-Channel signaling (ELAN network
interface). There is no restriction on the physical medium of the WAN.
Non-compressing G.711 codecs require 100BaseT Ethernet network
connectivity. A 10/100BaseT auto-sensing Ethernet network interface
routes the voice traffic from the IP trunk cards (TLAN subnet). Signaling
between cards and communication with the Optivity Telephony Manager TM
3.1 PC is transmitted over a 10BaseT Ethernet connection (ELAN subnet).
The application manages IP Trunk 3.01 (and later).
Figure 4 "IP Trunk 3.01 (and later) connectivity" (page 32) shows an IP
Trunk 3.01 (and later) configuration example.
Figure 4
IP Trunk 3.01 (and later) connectivity
In this document, TLAN subnet refers to the Telephony LAN subnet that
transmits the ITG voice and fax traffic. ELAN (Embedded LAN) subnet
refers to the management and signaling LAN subnet for the system site.
IP Trunk 3.01 (and later) depends on the managed IP network, not the
internet, because the managed IP network can provide adequate latency,
jitter, and packet loss performance to support VoIP with an acceptable voice
quality.
System requirements
The Media Card 32-port trunk card and the ITG-Pentium 24-port trunk cards
are able to reside in any of the following Meridian 1/CS 1000M systems
running CS 1000 Release 4.0 software:
•
Small Systems
•
Large Systems
IP Trunk 3.01 requires TM 3.1.
Customers must have the NTAK02BB (minimum vintage) SDI/DCH card
(Small Systems) or MSDL card (Large Systems) for ISDN Signaling
capability. If the customer does not have either of these cards, or does
not have an available DCH port on them, the customer must order these
cards to support ISDN functionality. Earlier vintages are not supported, as
the level of MCDN functionality required to support ITG-compatible ISL
is not available on earlier vintages.
Install a modem router on the ELAN subnet to provide remote support
access for IP Trunk 3.01 (and later) and other IP-enabled Nortel products.
The Nortel Netgear RM356 modem router integrates the functions of a
V.90 modem, a PPP remote access server, an IP router, and a 4-port
10BaseT Ethernet hub, and provides a range of security features that must
be configured to comply with the customer’s data network security policy.
The Netgear RM356 modem router can be ordered through many electronic
equipment retail outlets.
Table 2 "Software packages for Meridian 1/CS 1000M IP Trunk 3.01 (and
later)" (page 33) lists the required software packages.
Table 2
Software packages for Meridian 1/CS 1000M IP Trunk 3.01 (and later)
Hardware components for IP Trunk 3.01 (and later)35
Nortel Netgear RM356 Modem Router or equivalent is required for remote
support and must be ordered separately from retail outlets.
Inspect the IPE module to determine if it is equipped with non-removable
Molded Filter Connectors on the I/O Panel. For Large Systems
manufactured during the period of 1998-1999 and shipped in North
America, the IPE modules have the NT8D81BA Backplane to I/O Panel
ribbon cable assembly with a non-removable Molded Filter Connector. If
the TLAN subnet connection is 10BaseT use the NT8D81BA Backplane
to I/O Panel ribbon cable assembly, for a 100BaseT connection use the
NT8D81AA ribbon cable.
Table 3 "Hardware components for the Media Card 32-port trunk" (page 34)
lists the hardware components included in the IP Trunk 3.01 (and later)
packages for new installations.
Table 4 "Extra components for IP Trunk 3.01 (and later) trunk cards" (page
35) lists the extra components used by both the Media Card 32-port trunk
card and the ITG-Pentium 24-port trunk cards.
Table 4
Extra components for IP Trunk 3.01 (and later) trunk cards
Component
MSDL DCH cable (included in Large System package):
6 ftNTND26AA
18 ftNTND26AB
35 ftNTND26AC
50 ftNTND26AD
50 ft MSDL DCH Extender cableNTMF04AB
10 ft Inter cabinet cable NTCW84KA to SDI/DCH cableNTWE04AC
1 ft Intra cabinet cable NTCW84KA to SDI/DCH cableNTWE04AD
Shielded four-port SDI/DCH cable for the NTAK02BB SDI/DCH
card (included in Small System package)
PC Maintenance cable (for faceplate RS-232 maintenance port
to local terminal access)
Maintenance Extender cableNTAG81BA
Large Systems filter connector
50 pin I/O Panel Filter Connector Block with ITG specific
filtering for 100BaseTX (included in Large Systems package)
Backplane to I/O Panel ribbon cable assembly compatible with
NTCW84JA I/O Panel Filter Connector Block with ITG-specific
filtering for 100BaseTX TLAN subnet connection (replaces
NT8D81BA Backplane to I/O Panel ribbon cable assembly
equipped with non-removable Molded Filter Connectors)
Documentation
IP Trunk 3.01 (and later) NTP CD-ROM – MultilingualNTVQ61BA
PC Cards
C7LIU DCH PC Card with Layer 2 DCH SoftwareNTWE07AA
Product codes
NT8D81AA
Ordering rules and guidelines
IP Trunk 3.01 (and later) can be ordered as a VoIP trunk gateway with 32
ports, or as a software upgrade on an existing VoIP trunk gateway on the
Media Card 32-port trunk card or ITG-Pentium 24-port trunk card. One IP
Trunk card in the system must be equipped with a D-Channel PC Card kit.
One kit supports 12 Media Card 32-port trunk or 16 ITG-Pentium 24-port
trunk card with a maximum of 382 total ports.
Ordering rules for an IP Trunk 3.01 (and later) node
Initial configuration of an IP Trunk 3.01 (and later) node requires one
NTVQ01BB IP Trunk 3.01 Small and Large Systems 32-port package
with DCHIP as appropriate for the system. These packages include all
components needed for a single-card node, except for the cables that
provide interface to the MSDL and SDI/DCH cards. The following DCH
interface cables are included:
•NTND26AA (Large Systems)
•
NTAK19FB and NTWE04AD (Small Systems)
The following packages are required for IP Trunk 3.01 (and later):
•
ISDN Base (ISDN) package 145
•
ISDN Signaling Link (ISL) package 147
TM 3.1 is required and must be ordered separately.
For MSDL and DCHIP cards that reside in the same Large System UEM
equipment row, order:
•
NTND26 MSDL DCH cable in sufficient length to reach from the MSDL
to the I/O Panel of the IPE module that contains the DCHIP
For MSDL and DCHIP cards that reside in different Large System Universal
Equipment Modules (UEM) equipment rows in a multi-row Large System,
order:
•
NTMF04BA MSDL DCH Extender (50 ft.) cable to reach between the
I/O Panels of the two UEM equipment rows
For SDI/DCH and DCHIP cards that reside in different Small System
cabinets, order:
•
NTWE04AC Inter-cabinet cable (NTCW84KA to SDI/DCH cable-10 ft)
If IP trunk cards are being installed in IPE modules equipped with
NT8D81BA Backplane to I/O Panel ribbon cable assembly with Molded
Filter Connectors, on a 100BaseTX TLAN subnet connection, order:
•
NT8D81AA Backplane to I/O Panel ribbon cable assembly compatible
with NTCW84JA Filter Connector Block with ITG-specific filtering for
100BaseTX TLAN subnet connection
Inspect the IPE module to determine if it is equipped with Molded Filter
Connectors on the I/O Panel. Molded Filter Connectors were shipped in
North America during a period from 1998 to 1999. Molded Filter Connectors
can be used with 10BaseT TLAN subnet connections.
Ordering rules for IP Trunk 3.01 (and later) node expansion
To expand an IP Trunk 3.01 (and later) node, the following are required:
•
For each additional non-DCHIP card:
— one NTVQ92AA IP Trunk 3.01 (and later) Small and Large Systems
32-port expansion package (without DCHIP)
•For each additional DCHIP card:
— one IP Trunk 3.01 (and later) Small and Large Systems 32-port
package with DCHIP
Sparing ratios for IP Trunk 3.01 (and later) components
Sparing ratios for selected components are listed in Table 5 "Sparing ratios"
(page 37).
Table 5
Sparing ratios
ComponentSparing ratio
NTVQ92AA IP Trunk 3.01 (and later) Small and Large
Systems 32-port expansion package (without DCHIP) (for
repair only -- no RTU license)
"NTVQ91VA IP Trunk 3.01 (and later) Small and Large
Systems 32-port package with DCHIP
10:1
I/O cable assemblies
IP trunk card description
The Media Card 32-port trunk card and ITG-Pentium 24-port trunk card
provide a cost-effective solution for high-quality voice and fax transmission
over an IP network.
The IP Trunk cards are an IPE-based assembly designed for installation in
a Meridian 1/CS 1000M IPE shelf.
A Media Card 32-port trunk card occupies one slot and can have a
maximum of 32 ports. The ITG-Pentium 24-port trunk card is a two-slot
trunk card and can have a maximum of 24 ports. On the ITG-Pentium
24-port trunk card, a Peripheral Component Interconnect (PCI)-based
Digital Signal Processing (DSP) daughterboard provides voice processing
and supplies the packets to the IP Trunk 3.01 (and later) network using a
Pentium host processor. The Media Card 32-port trunk card has the DSP
connected to the main assembly. This main assembly is what compresses
speech into packets and supplies the packets to the IP Trunk 3.01 (and
later) network using an Intel StrongARM (SA) processor.
The IP trunk cards monitor the IP network for delay (latency) and packet
loss between other IP trunk cards. The card re-routes new calls to the
alternate circuit-switched trunk routes if the Quality of Service (QoS) of the
data network is not acceptable. Customers can configure QoS parameters
on the IP trunk cards to ensure that the IP Trunk 3.01 (and later) trunk route
is not used for new calls if the network QoS degrades below an acceptable
level. QoS monitoring is not available for Gatekeeper-routed endpoints
such as the CS 1000M.
20:1
8051 XAController firmware
The XAController firmware is delivered through the following formats:
•
ITGPFW57.BIN - 8051 XAController firmware for the ITG-Pentium
24-port trunk card
•
SMCFW67.BIN - 8051 XAController firmware for the Media Card 32-port
trunk card NTVQ01BA
Table 6 "Firmware compatibility matrix for the Media Card 32-port trunk
card" (page 39) gives the firmware compatibility for the Media Card 32-port
trunk card.
Table 6
Firmware compatibility matrix for the Media Card 32-port trunk card
NTVQ01BB Media Card 32-port trunk cards are factory-programmed with
Release 8.0 firmware. Any firmware feature upgrades are available on the
Nortel website.
Download this firmware from the Customer Support Software page. Go to
ttp://www.nortel.com. Follow the links to Customer Support and Software
h
Distribution or go to h
Card roles
The Media Card 32-port trunk card and ITG-Pentium 24-port trunk card can
have one or more of the following roles:
•
•
•
•
Firmware version
6.7
8.0
Not compatibleCompatible
ttp://www.nortel.com/support
Follower
Active Leader
Backup Leader
D-channel IP gateway (DCHIP)
NTVQ01BANTVQ01BB
CompatibleNot compatible
The card roles identify which systems are active systems/standby systems
and which are client systems. The Active Leader has a Node IP address on
the voice interface. This node IP is an alias IP which is added to the original
IP address on the voice interface. Other machines in the network use the
Node IP to keep track of the Active Leader.
Each Meridian 1/CS 1000M is usually configured with the following:
•
one IP trunk card that acts as an Active Leader
•
one IP trunk card that acts as a Backup Leader
•
at least one IP trunk card that provides DCHIP functionality
•
one or more IP trunk cards identified as Followers
In the TM 3.1 ITG application, the term Leader 0 refers to the IP trunk card
initially configured to perform the role of the Active Leader. The term Leader
1 refers to the IP trunk card that is initially configured to perform the role of
Backup Leader. The Active Leader and Backup Leader exchange the Node
IP address when the Active Leader goes out-of-service. The term Active
Leader indicates the Leader 0 or the Leader 1 card that is performing the
Active Leader role.
Leader 0 or Leader 1 can have Active Leader status. On system power-up,
Leader 0 normally functions as the Active Leader and Leader 1 as the
Backup Leader. At other times, the Leader card functions reverse, with
Leader 1 working as the Active Leader and Leader 0 working as the Backup
Leader.
The Leader, Backup Leader, Follower, and DCHIP cards communicate
through their ELAN network interfaces. For more information, see "Internet
Protocols and ports used by IP Trunk 3.01 (and later)" (page 131).
Follower
A Follower card is a Media Card 32-port trunk card and/or an ITG-Pentium
24-port trunk card which converts telephone signals into data packets and
data packets into telephone signals. For outgoing calls, Follower cards
provide dialed number-to-IP address translation.
Active Leader
The Active Leader card is an IP trunk card that acts as a point of contact for
all other Meridian 1/CS 1000 systems in the network.
The Active Leader card is responsible for the following:
•
distributing incoming H.323 calls to each registered Follower card in
its node and balancing load among the registered cards for incoming
IP calls
•
IP addresses for other cards in its node (see "Interactions among card
functions" (page 44))
•
serving as a time server for all IP trunk cards in its node
•
performing network monitoring for outgoing calls in its node
•
voice processing
All calls from a remote VoIP gateway node are first presented to the Active
Leader card. The Leader card maintains a resource table of all the IP trunk
cards in its node. The Active Leader card consults its internal IP trunk card
resource table to determine which card has the most idle channels and is
the least busy. Based on that information, the Active Leader card selects
the card to receive the new call.
In a multi-card IP Trunk 3.01 (and later) node, the Active Leader is busier
than the Follower cards. As a result, the channels on the Follower cards
are used first. Only after most of the channels on the Follower cards and
Backup Leader card are in use does the Active Leader card assign an
incoming call to itself.
After a channel on a card has been selected, the Active Leader sends a
message to the selected IP trunk card telling it to reserve a channel for the
new call. The Active Leader redirects the call to the selected IP trunk card.
All subsequent messages are sent directly from the remote VoIP gateway
node to the selected card.
Backup Leader
The Backup Leader card steps in when the Leader is out-of-service. This
minimizes service interruptions.
D-channel IP gateway
The ITG-Pentium 24-port or Media Card 32-port trunk card with D-channel
IP gateway (DCHIP) functionality (DCHIP card) is connected by the RS-422
cable to the Multi-purpose Serial Data Link (MSDL) card on the Meridian
1/CS 1000M Large Systems. It connects to the SDI/DCH Card on Small
Systems. The DCHIP Card is equipped with a DCH PC Card. The DCH PC
Card provides the RS-422 and LAPD functionality that is required for the
D-channel (DCH) interface to the system. The DCHIP Card is the network
side of the system ISL D-channel connection. The card is a tandem node in
the switch network, providing a single-to-multi-point interface between the
Meridian 1/CS 1000M and the IP Trunk 3.01 (and later) network. See Figure
The ISL connection to the Meridian 1/CS 1000M functions as it does in a
normal ISDN network. The ISL controls the call processing for calls over
analog ISDN Signaling Link (ISL) TIE trunks. With IP Trunk 3.01 (and later),
these ISL TIE trunks are located on the IP trunk cards. The IP Trunk 3.01
(and later) D-channel only controls IP trunk cards in the same IP Trunk
3.01 (and later) node. TM 3.1 administration relates the cards with trunks
to the DCHIP IP trunk card.
The IP trunk card uses ISDN messages for call control and communicates
with the Meridian 1/CS 1000M through the PC Card, using the RS-422 link.
On the Meridian 1, the MSDL provides the ISL DCH interface. The DCHIP
IP trunk card software performs the tandeming of DCH call control to the
H.323 protocol.
Each DCHIP trunk card can be associated with up to 382 trunks. The trunks
reside on all IP Trunk 3.01 (and later) IP trunk cards (ITG-Pentium 24-port
trunk cards and Media Card 32-port trunk cards) in the node. This creates a
functional grouping of IP trunk cards with the DCHIP trunk card providing
the DCH connectivity. If more than 382 trunks are required, additional
DCHIP trunk card groups are configured, each with a maximum of 382
related trunks. See Figure 6 "Leader, DCHIP, and trunks in an IP Trunk 3.01
Figure 6
Leader, DCHIP, and trunks in an IP Trunk 3.01 (and later) node
Card combinations
The Leader and DCHIP, or Follower and DCHIP, functions can reside on a
single IP trunk card or multiple IP trunk cards. If a Follower card is equipped
with a DCH PC card, it can function as a DCHIP trunk card. As an IP Trunk
3.01 (and later) node becomes larger with more trunk traffic, load balancing
should be configured. When load balancing is required, the Leader and
DCHIP functionality are placed on separate cards which are assigned
the least call traffic. For the largest IP Trunk 3.01 (and later) nodes and
networks, the Leader and DCHIP cards can be partially configured with
trunk ports or have no trunk ports at all.
An example configuration that allows for redundancy and backup is the
following:
To support more trunks, more DCHs can be added. Each DCHIP card can
support a maximum of 15 NT0961AA ITG-Pentium 24-Port Follower cards
or 11 NTVQ90BA Media Card 32-port Follower cards. This limit is due to
the maximum limit of 382 trunks in an ISL route.
Each DCHIP card controls a separate group of Follower cards. If a DCHIP
card fails, its associated Followers are removed from service as well. For
very large nodes, it is recommended that Follower cards be spread across
multiple DCHIPs, in order to provide some resiliency by allowing the IP Trunk
3.01 (and later) node to continue handling calls when one DCHIP card fails.
A DCHIP card and all of the IP trunk cards connected with it belong to one
Leader card. This means that the cards also belong to a single customer.
The group of IP trunk cards connected with one Leader is referred to as an
IP Trunk 3.01 (and later) node. If a single Meridian 1/CS 1000M system
has multiple customers requiring IP Trunk 3.01 (and later) connectivity,
a separate IP Trunk 3.01 (and later) node is required for each customer.
Multiple DCHIPs can be configured for each node.
All DCHIPs in an IP Trunk 3.01 (and later) node must be configured with the
same DCH protocol. If the user wants to use multiple DCH protocols, the
user must configure multiple IP Trunk 3.01 (and later) nodes.
Each customer requires one or more dedicated IP Trunk 3.01 (and later)
nodes. Trunks on the same IP Trunk 3.01 (and later) node share the same
dialing plan and IP network connectivity. IP Trunk 3.01 (and later) trunks
cannot be shared between customers that have independent numbering
plans and IP networks.
It is possible to configure multiple IP Trunk 3.01 (and later) nodes for one
customer. This configuration allows load balancing among multiple Leaders
for systems with more traffic than a single Leader card can support. The
configuration of multiple IP Trunk 3.01 (and later) nodes on one customer
requires splitting the dialing plan among the Leaders. Each Leader must
have a distinct range of the dialing plan. This restriction exists so that a
remote gateway can relate a DN with a single IP address.
For information about engineering an IP Trunk 3.01 (and later) node, refer to
"ITG engineering guidelines" (page 87).
Interactions among card functions
Active Leader and Follower card interaction
The Active Leader card controls the assignment of IP addresses for all new
ITG-Pentium 24-port and Media Card 32-port trunk cards in its node. If a
new IP trunk card is added as a Follower, the new Card Configuration data,
as programmed in TM 3.1, is downloaded only to the Active Leader card.
When it boots up, the new Follower card requests its IP address from the
Active Leader card through the bootp protocol. When the Follower cards
boot up, they receive their IP address and Active Leader card IP address
from the Active Leader card.
Follower cards continuously send Update messages to the Active Leader
card. These messages inform the Active Leader card of the Followers’ most
recent status and resources. The Active Leader sends Update messages
to the Follower cards, informing them of the updated dialing number to IP
address translation information. Also the Active Leader card continuously
sends messages about changes in the network performance of each
destination node in the dialing plan.
If a Follower card fails (for example, DSP failure), it reports to the Active
Leader that its failed resources are not available. The trunk ports involved
are considered faulty and appear busy to the Meridian 1/CS 1000M. Call
processing is maintained on the remaining IP Trunk 3.01 (and later) trunks.
If a Follower card loses communication with the Active Leader, all its ports
appear busy to the Meridian 1/CS 1000M. Alarms are raised by sending an
Simple Network Management Protocol (SNMP) trap to the IP addresses
in the SNMP manager list.
Active Leader and Backup Leader interaction
When a Leader card reboots into service, it sends bootp requests to
check whether an Active Leader card is present. If it receives a bootp
response, this indicates the presence of an Active Leader card and the
rebooting Leader becomes the Backup Leader. If it does not receive a bootp
response, this indicates the absence of an Active Leader and the rebooting
Leader becomes the Active Leader.
The Backup Leader monitors the heartbeat of the Active Leader by pinging
the Active Leader’s Node IP. In the event of the Active Leader’s failure (that
is, the Active Leader is not responding to the pinging of the Node IP address
by the Backup Leader), the Backup Leader takes over the Active Leader
role, in order to avoid service interruption. The Backup Leader assigns
the Node IP to its voice interface and announces its new status to all the
Follower cards. The Followers re-register with the new Active Leader and,
as a result, a new Resource Table is built immediately.
The Leader 0 and Leader 1 cards keep their node properties synchronized.
The Backup Leader receives a copy of the bootp.1 file, containing the bootp
table, from the Active Leader on bootup and when Node Properties are
downloaded to the Active Leader.
Critical synchronized data includes the following:
•
the card index:
— index 1 indicates Leader 0
— index 2 indicates Leader 1
— index 3 or greater indicates Follower
In the event of a Backup Leader failure, the Leader card generates an
SNMP trap to the TM 3.1 management station, indicating this failure.
If the Active Leader and Backup Leader are reset, removed, or disconnected
from the LAN at the same time, the entire IP Trunk 3.01 (and later) node is
put out-of-service. If this situation occurs, manual intervention is required to
recover the system.
Active Leader/Backup Leader and DCHIP card interaction
The Active Leader checks the status of the DCHIP card. The DCHIP card
must constantly inform the Leader of its DCH status and its card status.
When a DCHIP trunk card failure occurs, the associated trunks’ states
appear busy to the Meridian 1/CS 1000M, so the trunks will not be used for
calls. This blocks the normal software action of reverting to analog signaling
when an ISL DCH fails. If either end’s DCHIP or DCH connection fails, ISDN
protocol features across the IP network do not function. When a DCHIP
card fails, its associated Followers are also removed from service.
the Management MAC address (motherboard Ethernet address)
the Node IP address
the individual card IP addresses and card TNs for all IP trunk cards in
the IP Trunk 3.01 (and later) node
D-Channel number, card density and First CHID
In the case of a DCH failure, established calls are maintained; however, no
new calls can be made. Calls in a transient state are dropped.
ITG-Pentium 24-port trunk card (NT0961AA)
The ITG-Pentium 24-port trunk card was introduced as part of ITG Trunk
2.0. During the installation of the IP Trunk 3.01 loadware, the application
on the ITG-P 24-port card(s) (ITG-P) must be upgraded. It is essential to
ensure the latest software is loaded on the ITG-P card(s).
Description
The NT0961AA ITG-Pentium 24-port trunk card plugs into an Intelligent
Peripheral Equipment (IPE) shelf. Each ITG-Pentium 24-port trunk card
occupies two slots. ITG-Pentium 24-port trunk cards have a ELAN network
interface (10BaseT) and a TLAN network interface (10/100BaseT) on
the I/O panel. The ITG-Pentium 24-port trunk card has a DIN-8 serial
maintenance port connection on the faceplate and an alternative connection
to the same serial port on the I/O backplane.
Do not connect two maintenance terminals to both the faceplate and I/O
panel serial maintenance port connections at the same time.
The NT0961AA ITG-Pentium 24-port trunk card supports 24 ports per card.
The core ITG processor is an Intel Pentium II (266 Mhz).
The ITG-Pentium 24-port trunk card is responsible for converting the 64
kbit/s Pulse Code Modulation (PCM) speech from the DS-30X backplane
interface into packetized speech for transmission over the IP network. On
the daughterboard, the DSPs compress speech and feed the resulting
packets to the IP network.
Figure 7 "ITG-Pentium 24-port trunk card system connectivity and
messaging" (page 47) shows ITG-Pentium 24-port trunk card system
connectivity.The ITG card provides sufficient flexibility to emulate any
DS-30X signaling protocol. To support IP terminals, an ITG card emulates
the XDLC with attached Aries sets. Signaling on all DS-30 channels is
supported, allowing the ITG card to support up to 32 ports on a single card.
Figure 7
ITG-Pentium 24-port trunk card system connectivity and messaging
Faceplate indicators, controls, and interfaces
The NT0961AA ITG-Pentium 24-port trunk card has a double width
faceplate using the shortened lock latches, as shown in Figure 8 "NT0961AA
A single red, card status LED on the faceplate indicates the enabled/disabled
status of the 24 ports on the card. The LED is lit (red) during the power-up
or reset sequence. The LED remains lit until the card correctly boots and
assumes its role (that is, Leader, Backup Leader, Follower or DCHIP). If the
LED remains on, one of the following has occurred:
•
the self-test has failed (the Faceplate Maintenance Display indicates
the cause F:xx)
•the card has rebooted
•
the card is active, but there are no trunks configured on it (for example,
the card is a Leader or DCHIP)
•
the card is active and has trunks, but the trunks are disabled (that is, the
trunks must be enabled in LD 32)
During configuration, the error message "F:10" can appear. This error
indicates a missing Security Device. It occurs because Security Devices are
not implemented on ITG Trunk 2.0. Ignore this message.
See "ITG-Pentium 24-port trunk card faceplate maintenance display codes"
(page 423) for a complete list of faceplate codes.
Ethernet status LEDs
Ethernet status LEDs for the voice interface on the daughterboard display
the Ethernet activity as follows:
•Green is always on if the carrier (link pulse) is received from the TLAN
Ethernet hub.
•
Yellow flashes when there is data activity on the TLAN Ethernet hub.
•
During heavy traffic, yellow can stay continuously lit.
There are no Ethernet status LEDs for the ELAN network interface on the
motherboard.
Reset switch
A reset switch on the faceplate allows an operator to manually reset the
card without having to cycle power to the card. This switch is normally used
following a software upgrade to the card or, alternatively, to clear a fault
condition.
PC Card socket
There are two PC Card sockets. The faceplate socket accepts either a Type
I, a Type II, or a Type III PC Card and is designated ATA device A:. The
internal socket is reserved for the NTWE07AA C7LIU DCH PC Card on
the DCHIP.
Maintenance display
This is a four character, LED-based dot matrix display. It shows the card
boot sequence and is labeled with the card role as follows:
•
LDR = Active Leader
•
BLDR = Backup Leader
•
FLR = Follower
A properly-functioning IP trunk card displays one of the above codes. If an
IP trunk card encounters a problem, a fault code is displayed. For more
information, see "Media Card 32-port trunk card faceplate maintenance
The ITG-Pentium 24-port card has a DIN-8 (RS-232) maintenance port
(DCE) connection on the faceplate and an alternative connection to the
same serial port on the I/O backplane. Do not connect two maintenance
terminals to both the faceplate and I/O panel serial maintenance port
connections at the same time.
Ethernet TLAN network interface
The faceplate Ethernet TLAN network interface is a 9-pin, sub-miniature
D-type connector. The Ethernet TLAN network interface on the
daughterboard is identified as "lnPci1" in the ITG shell.
Backplane interfaces
The following interfaces are provided on the backplane connector:
WARNING
Do not connect a TLAN cable to the faceplate 9-pin Ethernet TLAN
network interface NWK. Connect the TLAN cable to the I/O cable.
DS-30X voice/signaling
This carries PCM voice and proprietary signaling on the IPE backplane
between the IP trunk card and the Intelligent Peripheral Equipment
Controller (XPEC).
Card LAN
This carries card polling and initialization messages on the IPE backplane
between the IP trunk card and the Intelligent Peripheral Equipment
Controller (XPEC).
RS-232 serial maintenance port
This provides an alternative connection to the serial maintenance port
that exists on the I/O backplane. Use the NTCW84KA or NTMF94EA I/O
panel breakout cable to access the port. A DIN-8 serial maintenance
port connection exists on the faceplate. Do not connect two maintenance
terminals to both the faceplate and I/O panel serial maintenance port
connections at the same time.
Assembly description
The ITG-Pentium 24-port trunk card assembly consists of a two-slot
motherboard/daughterboard combination, as shown in Figure 9 "Mechanical
assembly" (page 51). A PCI interconnect board connects the motherboard
The ITG-Pentium24-port trunk card is not user-serviceable. Figure
9 "Mechanical assembly" (page 51) is for information purposes
only. Do not remove the daughterboard from the motherboard.
Figure 9
Mechanical assembly
Media Card 32-port trunk card (NTVQ01BB) 51
Media Card 32-port trunk card (NTVQ01BB)
The NTVQ01BB Media Card 32-port trunk card provides a single slot
implementation in an IPE shelf for Large and Small Systems. During the
installation of the IP Trunk 3.01 loadware, the application on the Media
Card(s) must be upgraded. It is essential to ensure the latest software is
loaded on the Media Card(s).
Description
The Media Card 32-port trunk card is based on an integrated hardware
platform that delivers a single-slot ITG solution, with an increase in port
density from 24 ports to 32 ports. The Media Card 32-port trunk card
faceplate is shown in Figure 10 "Media Card 32-port trunk card" (page 52).
A single red LED indicates the enabled/disabled status of the card and
the status of the power-on self-test.
Where a DCHIP PC Card is installed in the Media Card 32-port trunk card
A:/ drive, the LED does not indicate the status of the DCHIP PC Card or
the DCHIP.
Release 6.7Release 8.0
14
1
Compact Flash Drive with lock
Pin Retention
Nil
Compact Flash Drive with
Retaining Clip
Reset button
The reset button enables the operator to manually reset the card without
cycling power to it. Use the reset button to reboot the card after a software
upgrade, or to clear a fault condition.
This slot (designated as Slot A:) accepts a Type I or II PC Card. It also
supports a DCHIP interface PC Card (D-Chip) to the system through the
NTMF29Bx cable.
Ethernet activity LEDs
The LEDs indicate 100BaseT, 10BaseT, and activity on both the ELAN
and TLAN network interfaces.
Maintenance display
The maintenance display is a 4-character LED-based dot-matrix display. It
displays the IP trunk card boot sequence and displays the card role as
follows:
•
•
•
A properly-functioning IP trunk card displays one of the above codes. If an
IP trunk card encounters a problem, a fault code is displayed. For more
information, see "Media Card 32-port trunk card faceplate maintenance
The Media Card 32-port trunk card has a DIN-8 (RS-232) maintenance port
(DCE) connection on the faceplate and an alternative connection to the
same serial port on the I/O backplane.
Backplane interfaces
The Media Card 32-port trunk card provides the following interfaces on the
backplane connector:
•
DS-30X voice/signalling
•
card LAN
•
one RS-232 serial COM port for the Command Line Interface (CLI)
•
ELAN 10BaseT and TLAN 10/100BaseT network interfaces
CAUTION
Service Interruption
Do not connect two maintenance terminals to both the faceplate
and I/O panel serial maintenance port connections at the same
time.
Use the following guidelines when installing the Media Card 32-port trunk
card:
•
Ensure CS 1000 Release 4.0 software is installed and running.
•
Ensure that the NTVQ01BB Media Card Firmware is version 8.0 (or
later)
•
Order the Alarm and Notification application package separately.
•For all MCDN features, the SDI/DCH NTAK02 card (Small Systems)
or the MSDL NT6D80 card (Large Systems) is required. These cards
must be ordered for each system.
•
For Large Systems which include the NT8D81AB moulded Tip/Ring
Backplane cable, replace it with the NT8D81AA non-moulded version
cable for 100BaseT operation. For more information on installation of
the new filter block, refer to "Install NTCW84JA Large System I/O Panel
50-Pin filter adapter" (page 194).
•
A security dongle and keycode mechanism are not required on the
Media Card 32-port trunk card.
Software delivery55
•
The new Option11C Cabinet door and grill (which allows more space
between the door and the cards) is required due to the space needed by
the DCHIP faceplate assembly. A cabinet upgrade kit, NTDK18AA, is
available for the following cabinets: NTAK11xC or earlier, and NTDK50.
•
A maximum of ten Media Card 32-port trunk cards can be installed in
a Large System cabinet for Class B compliance (EN55022:1998 and
EN55024:1998). There are no limitations on the number of Media Card
32-port trunk cards that can be installed in other Meridian 1/CS 1000M
systems.
Software delivery
The IP Trunk 3.01 software application is provided on the onboard
CompactFlash card for the Media Card 32-port trunk card.
A programmed CompactFlash (NTM405AB) card is shipped along with
every IP Trunk 3.01 system package card. The CompactFlash must be
installed on the Media Card 32-port trunk card.
IMPORTANT!
The software is downloadable from the Nortel website and is available to IP Trunk
customers free of charge.
trunk cards and older Media Card 32-port trunk cards can both be upgraded
as outlined in "Software upgrade" (page 59).
Replacing a CompactFlash PC Card (C:/ drive)
If it is necessary to remove the CompactFlash PC (CFlash) card, follow the
steps in Procedure 1 "Removing the CFlash card on NTVQ01BB" (page 57).
Then, follow the steps in Procedure 2 "Installing the CFlash card" (page
57) to install the new CFlash card.
WARNING
The Media Card 32-port trunk card does not require file transfers
to or from the A:/ drive for normal operation. If a CFlash ATA card
is to be used for file transfers to or from the A:/ drive, to C:/ drive,
Nortel recommends that the CFlash ATA card be formatted on the
Media Card 32-port trunk card before use.
CAUTION
Service Interruption
When replacing the CFlash, contact the Nortel Technical Support
Center.
CAUTION
CAUTION WITH ESDS DEVICES
Use ESDS precautions when handling the Media Card 32-port
trunk card.
Place the Media Card 32-port trunk card horizontally on a clean
bench.
Nortel Communication Server 1000
IP Trunk Fundamentals
NN43001-563 01.01 Standard
Release 5.0 30 May 2007
Page 58
58 System description
2
The metal clip should be pulled up and the new CFlash card should
be kept in the right position (see Figure 13 "CFlash card with metal
clip up" (page 58)).
3
Ensure that force is applied equally at both ends of the CFlash card
before pushing it in (see Figure 13 "CFlash card with metal clip up"
(page 58)).
Figure 13
CFlash card with metal clip up
4Gently insert the CFlash, so that the flash is fully in contact with the
connectors on the drive.
5
Push the metal clip down so that the CFlash is locked in (see Figure
14 "CFlash card with metal clip down" (page 58)).
Figure 14
CFlash card with metal clip down
—End—
WARNING
The Media Card 32-port trunk card requires the IP Trunk 3.01
(and later) application software (exec file) to be present on the C:/
drive (CFlash card) in order to run the IP Trunk 3.01 (and later)
application.
IP Trunk 3.01 (and later) software upgrades can be performed in three ways:
•
by FTP from TM 3.1
•
by FTP from the CLI
•
from a PC Card
The application (exec) file for the Media Card 32-port trunk card contains
a different CPU type definition from other IP trunk card types. When
performing an upgrade on an IP trunk node containing a mixture of Media
Card 32-port trunk cards, ITG-Pentium 24-port trunk cards, and ITG 8-port
trunk cards, each card type must be upgraded with its corresponding image
file. It is important that all cards in a node are using the same software
release, which means that a node upgraded to IP Trunk 3.01 (and later) can
no longer have an ITG 8-port trunk card in that node.
IMPORTANT!
IP Trunk 3.01 (and later) does not support the ITG 8-port trunk card.
Software upgrade 59
ATTENTION
Follow the steps in Procedure 3 "Upgrading IP Trunk 3.01 (and later)
software" (page 59) to upgrade to IP Trunk 3.01 (and later) software.
Procedure 3
Upgrading IP Trunk 3.01 (and later) software
StepAction
1
Download the latest software upgrade information from the
Nortel website to the TM 3.1 PC or to an FTP server. Go to
h
ttp://www.nortel.com. Follow the links to Customer Support and
Software Distribution or go to http://www.nortel.com/support.
2
See "Check and download IP trunk card software in OTM 2.1 (and
later)" (page 271) for information on how to upgrade the software by
FTP from TM 3.1.
See "Transfer files through the Command Line Interface" (page 375)
and "Upgrade IP trunk card software using FTP" (page 377) for
information on how to upgrade the software by FTP from the CLI.
A CompactFlash PC Card containing the latest software version can
be obtained from Nortel. See "Upgrade IP trunk card software by PC
Card" (page 378) for information on how to perform the upgrade.
When the upgrade file has been downloaded, install the new IP
Trunk 3.01 (and later) application software onto the IP trunk card.
Follow the application software upgrade procedure as described in
Nortel Communication Server 1000
IP Trunk Fundamentals
NN43001-563 01.01 Standard
Release 5.0 30 May 2007
Page 60
60 System description
"Transmit card properties and dialing plan" (page 417) or in "Transfer
files through the Command Line Interface" (page 375).
—End—
Media Card application identification labels
Media Card application identification labels (see Figure 15 "Media Card
identification labels" (page 60)) are provided with every Media Card 32-port
trunk card package. Affix the appropriate label to the Media Card’s faceplate
(see Figure 16 "Labeled Media Card" (page 60)).
Figure 15
Media Card identification labels
Figure 16
Labeled Media Card
Interoperability with earlier versions of ITG Trunk
When Media Card 32-port trunk cards are implemented in existing networks
with nodes comprised of ITG Trunk 2.xx, Release 19 or earlier, fax calls do
not work because of protocol incompatibility. Voice calls between ITG Trunk
2.1 and ITG Trunk 2.0 or ITG Trunk 1.0 operate without restrictions.
If an upgrade from ITG Trunk 2.xx, Release 19 or earlier, is projected to
take several days and fax support is needed during this time, first upgrade
the individual nodes to ITG Trunk 2.xx Release 23. When the network
is upgraded to ITG Trunk 2.xx Release 23, upgrade again to the latest
software release. The interim upgrade step is only required if fax support is
needed during the upgrade process.
When the Media Card 32-port trunk cards are upgraded to or installed
with IP Trunk 3.01 (and later), fax calls do not work to nodes running ITG
Trunk 2.xx Release 19 or earlier. This limitation is due to the same protocol
incompatibility that exists between ITG Trunk 2.1 and ITG Trunk 2.xx and
earlier.
Fax Tone Detection Configuration
For IP Trunk 3.01 (and later) fax operation, the V.21 Tone detection check
box must be selected in TM 3.1 in the Configuration window, under the DSP
profile tab. For more information, see "Configure DSP profiles for the IP
Trunk 3.01 (and later) node" (page 242).
ISDN Signaling Link
ISDN Signaling Link (ISL) provides the capability of replacing conventional
analog trunk signaling with out-of-band ISDN D-channel signaling.
ISDN Signaling Link 61
The ISL interface makes available the flexibility of using ISDN signaling to
analog facilities. When no Primary Rate Interface (PRI) exists between
two Meridian 1/CS 1000M systems, ISL operates in dedicated mode.
A dedicated point-to-point signaling link is established between the two
systems. The signaling information for the selected analog trunks is
transported over the ISDN signaling link. The analog ISL TIE trunks are for
user voice transport. If the D-channel link is down, call control returns to
normal in-band analog trunk signaling.
The ITG is similar to the existing ISL configuration where there is a Virtual
Private Network (VPN) between Meridian 1/CS 1000M systems. Instead of a
one-to-one connection, multiple switches can be networked through a single
ISL interface at each site. Figure 17 "ITG configuration" (page 62) shows an
IP Trunk 3.01 (and later) trunk configuration with three Meridian 1/CS 1000M
systems. The IP Trunk 3.01 (and later) trunk simulates an analog facility.
The ISL interface is connected to a DCHIP PC Card which provides ISDN to
VoIP tandeming. All IP Trunk 3.01 (and later) IP trunk cards (DCHIP, Leader,
and Follower) are connected through the ELAN subnet. The IP trunk cards
communicate with remote switches through the IP network.
ISDN signaling between the Meridian 1 and IP Trunk 3.01 (and later)
supports the delivery of Calling Line Identification (CLID) and feature
messaging. ISL DCH signaling provides the necessary signaling connection
over which data, including CLID and feature-specific messaging, can be
passed.
On Large Systems, the DCH interface to the Meridian 1/CS 1000M uses the
MCDN or QSIG GF protocols and their variants to transmit call and feature
control messages to the DCHIP card. Small Systems use only MCDN
because the NTAK02BB SDI/DCH card does not support QSIG protocols
for ISL. The DCH interface uses these protocols and their variants, as they
have the following advantages:
•
ISL configuration support
•
symmetry (incoming and outgoing call messaging is the same)
•
near H.323 standard
QSIG GF Name Display is the only supported QSIG supplementary service.
The ITG feature complies with H.323 Basic Call Q.931 signaling. This part
of the H.323 standard (H.225) defines the messaging used to setup and
release basic calls. A mechanism is implemented to enable the passing
of ISDN messaging through the IP network between the two endpoints.
The call is set up using the H.323 standard signaling with encapsulated
ISDN-specific information. This mechanism allows interworkings with other
gateways.
The DCHIP card provides the tandem between the ISDN signaling and the
H.323 protocol. If the DCHIP functionality is combined with the Follower
card, messages are sent between the DCH Processor and the H.323
Processor. Most configurations split this functionality between the DCHIP
and Followercards. Figure 18 "Signal flow from the DCH to the H.323 stack"
(page 63) shows the signal flow from the DCH to the H.323 stack.
Figure 18
Signal flow from the DCH to the H.323 stack
For further information on ISDN Signaling Link (ISL), refer to System
Management Reference (NN43001-600), ISDN Primary Rate Interface
Installation and Commissioning (NN43001-301), and ISDN Primary Rate
Interface Maintenance (NN43001-717).
Inter-card signaling paths
The Leader, DCHIP, and Follower cards communicate using their ELAN
network interface IP addresses. Figure 19 "IP Trunk 3.01 (and later) card
signaling paths" (page 64) illustrates the IP signaling paths used inter-card
and between the cards and the system in the ITG offering.
Figure 19
IP Trunk 3.01 (and later) card signaling paths
In Figure 19 "IP Trunk 3.01 (and later) card signaling paths" (page 64),
the DS-30X connection is part of the IPE shelf’s backplane. The ISL DCH
connection is a cable that runs from the "octopus" breakout cable, on
the back of the IPE cabinet, to one of the MSDL’s RS-422 ports. The
Leader/Follower card messages normally travel over the TLAN subnet. The
DCHIP messages travel over the ELAN subnet - a 10BaseT LAN connected
to each IP trunk card and the TM 3.1 PC. A separate 10/100BaseT LAN
transmits the voice/fax data to the remote VoIP systems.
Dialing plans
Dialing plan configuration allows customers to set up routing tables to route
calls to the appropriate destination, based on dialed digits. The dialing plan
is configured through the Electronic Switched Network (ESN) feature, using
TM 3.1 or overlays in the system. With ESN configuration, the system can
route outgoing calls to the IP trunk card. Address translation allows the IP
trunk card call processing to translate the called party number to the IP
address of the terminating IP Trunk 3.01 (and later) node and to deliver calls
to the destination through the IP network.
The ITG-Pentium 24-port and Media Card 32-port trunk cards support the
following dialing plans:
•North American dialing plan
•
Flexible Numbering Plan
Customer-defined Basic Automatic Route Selection (BARS) and Network
Alternate Route Selection (NARS) Access Codes are used to access the
dialing plans.
The IP Trunk 3.01 (and later) dialing plan supports a single customer per
IP Trunk 3.01 (and later) node and multiple IP Trunk 3.01 (and later) nodes
per Meridian 1/CS 1000M system. A customer can have multiple IP Trunk
3.01 (and later) nodes in a system, but each node can only support the
dialing plan of a single customer. Multiple customers will require multiple
nodes per system.
Multi-node configuration
The following example explains a possible configuration between two
Meridian 1/CS 1000M switches to achieveboth resiliency into the IP network
and load balancing.
Meridian 1/CS 1000M switch A has two IP Trunk 3.01 (and later) nodes, A1
and A2, for the destination NPA 613. A Route List Block (RLB) is created, in
order to have two route entries (one for each IP Trunk 3.01 (and later) node).
If the trunks of node A1 are all in use or node A1 is down, call traffic is routed
to node A2. This provides resiliency by preventing failure of a single IP Trunk
3.01 (and later) node (for example, DCH failure or Leader subnet fails) from
completely eliminating VoIP service for a Meridian 1/CS 1000M system.
It is desirable to distribute calls to multiple nodes at a remote destination
Meridian 1/CS 1000M. The configuration of multiple dialing plan entries at
the local IP Trunk 3.01 (and later) node allows routing based on the dialed
digits.
For example, Meridian 1/CS 1000M switch B node B1 has two entries for
NPA 408 and 4085, which point to nodes A1 and A2 of Meridian 1/CS 1000M
switch A, respectively. Calls from B1 with dialed digits 408-5xx-xxxx are
routed to the IP Trunk 3.01 (and later) node A1 while all other 408-xxx-xxxx
calls are routed to IP Trunk 3.01 (and later) node A2.
The North American dialing plan is used to make public network calls
through the private IP network. However, calls are not directly routed to the
Central Office (CO) through the LAN connection. Instead, a tandem switch
with voice trunk connections, including T1 ISDN PRI, serves as the gateway
to route voice calls coming through the LAN to the voice trunk.
Figure 20 "North American dialing plan — call flow" (page 66) shows
DN 7000 placing a public call, through the private LAN, by dialing
1-415-456-1234 or 566-1234. The IP trunk card with IP address
47.82.32.124 searches for the Numbering Plan Area (NPA) or Local
Exchange Code (NXX) tables with the matched NPA or NXX entries. When
an entry is found, the corresponding IP address is used to send H.323 call
setup messages to the gateway (a Meridian 1/CS 1000M with an IP address
of 47.82.32.123), which routes the call to the PSTN through a regular CO
or DID trunk.
The translation table is expanded to allow extended, three-to six-digit NPA
codes. For example, DNs, such as 1-415-456-XXXX and 1-415-940-XXXX,
can have different destination IP addresses.
Figure 20
North American dialing plan call flow
Flexible Numbering Plan
A Flexible Numbering Plan (FNP) allows the length of Location Codes
(LOCs) to vary from node to node. As well, the total number of digits dialed
to reach a station can vary from station to station. It also allows flexibility for
the length of the location codes from node to node. An FNP can be used
to support country-specific dialing plans. FNP also allows users to dial
numbers of varying lengths to terminate at a destination. Flexibility of the
number of digits which can be dialed is achieved using Special Numbers
(SPNs).
IP Trunk 3.01 (and later) and ITG Trunk 2.x support a mixed network of
remote nodes with ESN5 and standard (that is, non-network) signaling.
ESN5 is an extension of MCDN signaling which can be used by IP Trunk
3.01 (and later), ITG Trunk 2.x, and IP Peer ( CS 1000M).
ESN5 inserts the Network Class of Service (NCOS) prefix ahead of the
dialed numbers. Make sure that, if ESN5 is to be used, it is provisioned on
both the IP trunk cards and the Route Data Block (RDB) for that node. If
ESN5 is provisioned for an IP Trunk 3.01 (and later) node, all remote ITG
2.x and IP Trunk 3.01 (and later) node must have that node provisioned as
"SL1ESN5" in the dialing plan. If this is not done, a default NCOS is inserted
by the ESN5 node receiving the call from the non-ESN5 VoIP gateway. Fore
more information, see "ESN5 network signaling" (page 229).
Echo cancellation
All telephony voice services now in use reflect some level of echo back
to the user. The term "echo" refers to the return of a signal’s reflection
to the originator.
Dialing plans67
Packet voice networks introduce sufficient latency to cause what a caller
would consider an audible echo. The echo path is round-trip. Any speech
coding, packetization, and buffering delays accumulate in both directions of
transmission, increasing the likelihood of audibility.
Echo cancellation reduces feedback sounds and background noise for
clearer voice quality. Some less advanced IP telephony products do not
include echo cancellation circuitry, resulting in voice quality of a level below
business communications standards. Without echo cancellation, the talking
parties can hear varying levels of echo as they speak.
Echo canceller tail delay
Early versions of ITG Trunk DSPs and DSP firmware had a maximum echo
canceller (ECAN) tail delay of 32 ms. More recent cards and firmware
support higher tail delays, with the ITG-P and the Media Card 32-port card
supporting up to 128 ms. However, when the capability was added, the
default in TM 3.1 remained unchanged at 32 ms, even though the ECAN
performance was significantly better with 128 ms. This problem has been
resolved in TM 3.1, but ITG Trunk and IP Trunk nodes defined by customers
with the original TM 3.1 software still use the incorrect default value.
Recent releases of TM 3.1 that are properly configured, with all applicable
patches and the fix integrated, have the default for new systems set to 128
ms. This results in all new nodes being given the correct default value.
However, it will not change the value on systems that are already configured
unless the user deliberately changes the value.
IP Trunk 3.01 includes an enhancement to accommodate this issue. Since
a 32-ms ECAN tail delay is usually only provisioned "by default" and not
by deliberate user programming, the IP Trunk 3.01 application maps an
ECAN tail delay of 32 seconds to the corrected default of 128 ms. This
addresses the vast majority of users who want the optimum available ECAN
performance. However, a small number of users, for various reasons, may
want the 32-ms tail delay.
Users that can accept poorer echo performance and really want a 32 ms
delay can use a value of 8 ms, which the IP Trunk application maps to 32
ms. A delay of 8 ms is completely unacceptable to end users, so this does
not result in any loss of user capabilities. In addition, a value of 16 ms,
which is also unsatisfactory, is mapped to a delay of 64 ms, maintaining the
same two-to-one ratio with the next lower value in both the TM 3.1 and IP
Trunk environment. (In this case, the 8 ms value is half the 16 ms, and the
32 ms value is half the 64 ms value.)
Table 8 "Echo canceller tail delay mapping from OTM to IP Trunk 3.01"
(page 68) shows the mapping between the delay value configured in TM
3.1 and the actual delay value used in IP Trunk 3.01. The actual configured
delay value can be displayed using the CLI command itgCardShow.If
the TM 3.1 value is mapped, "Default - xxx" is displayed, where "xxx" is
the mapped value. If the TM 3.1 value is 64, 96, or 128 ms, "Value from
TM 3.1 - xxx" is displayed.
Table 8
Echo canceller tail delay mapping from TM 3.1 to IP Trunk 3.01
Provisioned in TM 3.1 (in ms)Value used by IP Trunk 3.01
832
1664
32 (default value in IP Trunk 3.01 and
earlier)
6464
9696
128 (default value in IP Trunk 3.01)128 (default value in IP Trunk 3.01)
Speech activity detection reduces the IP bandwidth used by typical
voice conversations. When Speech Activity Detection is enabled, no
voice samples are sent during periods of silence (from one side of the
conversation or the other). When a caller stops speaking, instead of a
"dead" line, the listener hears "comfort noise" generated to match the
previous background noise level when the caller was speaking.
Coders can send silence frames before the end of transmission during a
period of silence. Coders might omit sending audio signals during periods of
silence after sending a single frame of silence, or send silence background
fill frames, if these techniques are specified by the audio codec in use.
This background white noise keeps the telephone from sounding like the
line has gone dead - the listener can tell that the call is still up, and that
the person at the other end has merely stopped speaking. This technique
allows pauses during calls to sound almost the same as they would on a
standard telephone line. The primary benefit of Speech activity detection is
that it allows the IP Trunk to use bandwidth only when it needs to send voice
samples, thereby saving expensive WAN bandwidth for data traffic or other
voice and fax calls. Since normal telephone conversations include pauses,
and only one side is normally speaking, Speech activity detection reduces
the bandwidth used on a call by more than half.
Dialing plans69
For applications that send no packets during silence, the first packet after
a silence period is distinguished by setting a marker bit in the Real Time
Protocol (RTP) data header. Applications without Speech Activity Detection
set the bit to zero.
DTMF Through Dial
Preservation and transport of tones through the IP Trunk 3.01 (and later)
network is critical for Interactive Voice Response (IVR) services. IP Trunk
3.01 (and later) can be configured to ensure that DTMF tone information is
included in the packets that are sent through the IP Trunk 3.01 (and later)
network and that the tones are re-transmitted by the far-end gateway. The
duration information for DTMF signals is not transmitted; that is, long DTMF
bursts are reduced to a short standard duration.
Callers can access traditional Voice Mail or IVR services (for example,
"Press 1 for more information" or "Press 2 to be connected to our customer
service department"). Services that depend on long DTMF bursts cannot
be accessed.
In order to ensure that DTMF tones are being transmitted properly, the DSP
must be configured correctly in TM 3.1. If the IP Trunk 3.01 (and later)
node is configured to use a voice codec other than G.711, "DTMF Tone
Detection" must be selected (checked) in TM 3.1. See Figure 21 "DTMF
tone detection" (page 70). For more information on how to configure the
IP Trunk 3.01 (and later) DSP, see "Configure DSP profiles for the IP Trunk
3.01 (and later) node" (page 242). If the IP Trunk 3.01 (and later) node is
using G.711 without "DTMF Tone Detection" checked, there is no guarantee
that DTMF tones will be properly transmitted to the far end, due to the
possibility of latency or packet loss.
Figure 21
DTMF tone detection
Quality of Service
Quality of Service (QoS) is the gauge of quality of the IP network between
two nodes. As QoS degrades, existing calls suffer from poor voice and
fax quality. New calls will not be initiated if transmissions degrade below
an acceptable level.
Behavioral characteristics of the IP network depend on the following:
The Type of Service (ToS) bits in the IP packet header can affect how
efficiently data is routed through the network. For further information on
ToS, see "Type of Service" (page 76).
Packet jitter related to latency affects the quality of real-time IP
transmissions. For good voice quality, the IP trunk card reassembles the
voice packets in an ordered continuous speech stream and plays them out
at regular intervals despite varying packet arrival times.
The user configures a required QoS for the IP Trunk 3.01 (and later) node in
TM 3.1. The QoS value determines when calls fallback to alternate facilities
due to poor performance of the data network. The QoS value is between 0.0
and 5.0, where 0.0 means never fallback to alternate facilities and 5 means
fallback to alternate facilities unless the voice quality is perfect. When the
QoS for outgoing calls, as measured by the Leader card, falls below the
configured value, calls fallback to alternate facilities. Once the QoS rises
above the configured value, all new outgoing calls are routed through the
IP network.
QoS is measured for each remote gateway. For example, if a given Leader
has three remote leaders in its dialing plan table, it performs three QoS
measurements and calculations (one per remote gateway).
Since IP trunks use the same port for both voice and fax, the same QoS
thresholds apply for both voice and fax calls. Network requirements for fax
are more stringent than for voice. Fax protocols, such as T.30, are more
sensitive to transmission errors than the human ear.
Quality of Service parameters
Quality of Service for both voice and fax depends on end-to-end network
performance and available bandwidth. A number of parameters determine
the ITG voice QoS over the data network.
Packet loss
Packet loss is the percentage of packets sent that do not arrive at their
destination. Packet loss is caused by transmission equipment problems
and congestion. Packet loss can also occur when packet delays exceed
configured limits and the packets are discarded. In a voice conversation,
packet loss is heard as gaps in the conversation. Some packet loss, less
than five percent, can be acceptable without too much degradation in
voice quality. Sporadic loss of small packets can be more acceptable than
infrequent loss of large packets.
Packet delay is the time between when a packet is sent and when it is
received. The total packet delay time consists of fixed and variable delay.
Variable delay is more manageable than fixed delay, as fixed delay is
dependent on network technology. Variable delay is caused by the network
routing of packets. The IP Trunk 3.01 (and later) node must be as close
as possible to the network backbone (WAN) with a minimum number of
hops, in order to minimize packet delay and increase voice quality. ITG
provides echo cancellation, so that a one-way delay up to 200 milliseconds
is acceptable. For more information about Echo Cancellation, see "Echo
cancellation" (page 67).
Delay variation (jitter)
The amount of variation in packet delay is referred to as delay variation or
jitter. Jitter affects the ability of the receiving IP trunk card to assemble
voice packets into a continuous stream when the packets are received at
irregular intervals.
Latency
Latency is the amount of time it takes for a discrete event to occur.
Bandwidth
Bandwidth is a measure of information carrying capacity available for a
transmission medium. The greater the bandwidth the more information
that can be sent in a given amount of time. Bandwidth is expressed in bits
per second (bps).
Network performance utilities
Two common network performance utilities, Packet InterNet Groper (PING)
and Traceroute, are described in this section. Other utilities can be used to
gather information about IP Trunk 3.01 (and later) network performance.
These descriptions are for reference purposes only. Traceroute is not part of
the IP Trunk 3.01 (and later) product.
Because network conditions can vary over time, collect performance data
over a period of at least four hours. Use performance utilities to measure
network performance from each IP Trunk 3.01 (and later) node to every
other IP Trunk 3.01 (and later) node in the network.
Packet InterNet Groper (PING)
Packet InterNet Groper (PING) sends an Internet Control Message Protocol
(ICMP) echo request message to a host, expecting an ICMP echo reply.
This allows the measurement of the round-trip time to a selected host. By
sending repeated ICMP echo request messages, the percentage of packet
loss for a route can be measured.
Traceroute uses the IP Time-To-Live (TTL) field to forward router hops to
a specific IP address. A router must not forward an IP packet with a TTL
field of 0 or 1. It must, instead, discard the packet and return an ICMP
"time exceeded" message to the originating IP address. Traceroute uses
this mechanism by sending an IP datagram with a TTL of 1 to the specified
destination host. The first router to handle the datagram returns a "time
exceeded" message. This identifies the first router on the route. Traceroute
sends out a datagram with a TTL of 2. This causes the second router on the
route to return a "time exceeded" message, and so on, until all hops have
been identified. The Traceroute IP datagram has a port number unlikely to
be in use at the destination (usually >30,000). This causes the destination
to return a "port unreachable" ICMP packet which identifies the destination
host. Traceroute can be used to measure round-trip times to all hops along
a route, identifying bottlenecks in the network.
IP Trunk 3.01 (and later) uses the E-Model, a method similar to the ITU-T
Recommendation G.107, to determine voice quality. This model evaluates
the end-to-end network transmission performance and outputs a scalar
rating, R, for the network transmission quality. IP Trunk 3.01 (and later)
uses a simplified version of the model to correlate the network QoS to the
subjective Mean Opinion Score (MOS).
MOS is a numerical scale used to rate voice quality. When MOS is equal to
5.0, voice quality is good. When MOS is equal to 0.0, voice quality is bad.
For packet loss over 16%, the MOS value is set to 0, and the remote node is
considered to be in fallback mode.
End-to-end latency
IP Trunk 3.01 (and later) network end-to-end latency consists of several
components: routing delay on the IP Trunk 3.01 (and later) network, frame
duration delay and jitter buffer delay on the codec, and delay on the
circuit-switched network. The determination of end-to-end delay depends
on the dynamics of the IP Trunk 3.01 (and later) network and the detailed
service specification.
MOS values are calculated based on the routing delay and frame duration
and jitter buffer delay on the codec. These latencies must be taken into
consideration during the engineering of the total network’s latency. If the
end-to-end latency of the network is specified and the latency of the PSTN
circuit-switched components is removed, the remainder is the latency
available for the IP trunks. This latency value plays a large role when
configuring IP Trunk 3.01 (and later) node QoS values in TM 3.1.
For instance, assume the end-to-end network latency is 300 milliseconds
(ms) and the part of that latency which the IP network contributes is 180 ms.
Furthermore, assume the network has low packet loss. Using the G.711
codec, this means the configured QoS can be a minimum of 4.3. If the
latency in the IP network increases, the configured QoS is not met and
fallback to alternate facilities occurs.
Equipment Impairment factor
Equipment Impairment factors are important parameters used for
transmission planning purposes. They are applicable for the E-Model.
For information on QoS engineering guidelines, refer to "ITG engineering
guidelines" (page 87).
Fallback to alternate facilities
IP Trunk 3.01 (and later) continuously monitors and analyzes QoS data.
When IP Trunk 3.01 (and later) detects IP network congestion, and the
QoS is below a pre-defined value, new calls routed to the remote gateway
are rejected. Instead, the Meridian 1/CS 1000M routes them over non-IP
facilities. The Stepback on Congestion over ISDN feature provides fallback
to alternate facilities functionality.
Triggering fallback to alternate trunk facilities
A key background activity of IP Trunk 3.01 (and later) is to monitor the
network’s QoS between itself and each remote IP gateway configured in
the dialing plan. When the QoS is below the defined acceptable level for a
given IP Trunk 3.01 (and later) destination node, all outgoing calls from the
near-end Meridian 1/CS 1000M to the far end Leader are re-routed through
alternate circuit-switched trunk facilities. All calls that the switch is trying to
set up are re-rerouted; established calls cannot fall back.
The Meridian 1/CS 1000M provides alternate routing based on BARS or
NARS. BARS/NARS translates the dialed LOC, NPA, NXX, or Special
Number (SPN) into an entry on the Route List Block (RLB) and searches
the trunks in the associated Route Data Block (RDB).
The trigger for fallback to alternate trunk facilities is defined per call, per
customer. The local Active Leader makes the decision to use the fallback
feature. The selection of routes is based on the customer-configured
database. The customer must configure the alternate routing to the PSTN
in the Meridian 1/CS 1000M database.
The fallback to alternate facilities uses an ISDN DCH mechanism. The
Step Back on Congestion over ISDN feature provides fallback to alternate
trunk facilities functionality. When the Meridian 1/CS 1000M presents an
outgoing call and receives a release message back that indicates network
problems, Stepback on Congestion allows a new route to be found for the
call (for instance, the PSTN). The route selected depends on the customer’s
database. If an alternate route is not configured in the route list, the calls
rejected by IP Trunk 3.01 (and later) is routed to some other treatment.
Fallback is optional, based on the configuration of the route list.
Figure 22 "Example of a fallback to alternate facilities situation" (page 75)
shows the fallback to alternate facilities functionality.
Figure 22
Example of a fallback to alternate facilities situation
Fallback to alternate facilities75
Fallback in IP Trunk 3.01 (and later)
In QoS monitoring, the local node queries the remote node and gets a
response; the remote node queries the local node and gets a response.
If the remote node cannot query the local node, QoS monitoring is not
available. When an IP Trunk 3.01 (and later) node uses a Gatekeeper to
resolve an address, IP Trunk 3.01 (and later) cannot monitor QoS and
provide fallback. This function resides with the device resolving the address.
As a result, for all calls going to the Gatekeeper, such as in IP Peer
Networking, no fallback can occur. The call either goes through with
possibly a lower QoS, or the call clears instead of falling back. All QoS
control is in the hands of the Gatekeeper.
However, for calls using the ATPM static address tables, the IP Trunk 3.01
(and later) Leader retains awareness of network status and can cause
fallback to the PBX, if needed.
The full QoS fallback function is available for locally provisioned addresses.
IP Peer and Qos
The IP Peer Networking nodes do not support QoS monitoring. The
capability must be enabled for both sides in order for it to work, but it
cannot be enabled for IP Peer Networking. Therefore, do not enable QoS
monitoring for any numbers terminating on an IP Peer Networking node. If
this is done, the IP Peer Networking node is unreachable for that IP Trunk
3.01 (and later) node.
IP Trunk 3.01 (and later) nodes can perform QoS monitoring only on remote
IP Trunk 3.01 (and later) nodes provisioned locally with SL1, SL1 with ESN5
node capabilities.
Return to the IP network
Unless the DCH is down and all trunks appear busy to the system, outgoing
calls are introduced to the IP Trunk 3.01 (and later) node. Each call is tested
against the outgoing address translation and Quality of Service (QoS)
for the destination node. After the QoS returns to an acceptable level,
all new outgoing calls are again routed through the IP network. The call
connections that were established under the fallback to alternate facilities
condition are not affected.
Type of Service
The IP packet handler has a byte of data for Type of Service (ToS). This
byte allows the user to indicate a packet’s priority so that routers can more
efficiently handle data packets. For example, a router can decide to queue
low priority data while immediately passing packets marked as high priority.
The TM 3.1 User Interface allows two ToS values to be configured: data and
control. Data packets transmit the voice or fax call’s data, while control
packets setup and maintain the call. Both can be configured for any value
in the range of 0 – 255 (0 is the default). When an IP Trunk 3.01 (and
later) node is configured, ToS bits are initially set to default values. The
TM 3.1 IP Trunk 3.01 (and later) node administration interface allows the
customer to configure these bits for potentially better interworking with
different manufacturers’ routing equipment. The extent of any improvement
from setting these ToS bits depends on the network routing equipment.
Improvements can vary depending on the router’s prioritization algorithms.
The data ToS is placed in every voice or fax data packet sent from the IP
trunk card. To optimize the speech quality, ToS is usually configured for
low-latency and high-priority.
The control ToS is placed in every signaling message packet sent from the
IP trunk card. Signaling links use Transmission Control Protocol (TCP)
which provides a retransmission mechanism. In addition, the latency of the
control packets is not as critical as it is for the data packets.
Each entry in the routing table has a configurable ToS. ToS values are
configured in the DSP Profile window. For a route entry to be selected for
an outgoing packet, both the configured route and the ToS must match. Two
cases must be considered: local subnet traffic and remote traffic.
The remote subnet packets is the H.323 call data for an IP Trunk 3.01 (and
later) node which is not on the local subnet and must go through a router.
There is a default gateway entry (0.0.0.0) that specifies the gateway address
for this traffic. The ToS does not matter for this route. If the route and ToS
do not match any of the other route entries, the packet is routed here. The
entry is configured for the TLAN network interface.
Local subnet packets is the H.323 call data intended for another IP Trunk
3.01 (and later) node connected to the same subnet. This can be the
immediate subnet. For traffic to be sent on the local subnet, the routing
table entry for the TLAN network interface must be selected. Each table
entry (except the default route) has a ToS value configured against it. Since
there are two ToS values configured (one for control data and one for voice
data), there must be two route entries for the local subnet in the table.
If both table entries are not present, a condition occurs where packets for
voice, control, or both can be sent to the default route because the ToS
does not match the local subnet entry. These packets go to the router and
then back on the subnet, wasting router resources and increasing traffic on
the subnet.
The IP trunk card configures two route table entries for the local subnet if a
different ToS is configured for the voice and control packets. Otherwise, a
single entry is created.
CAUTION
Service Interruption
Only technical personnel with detailed knowledge of router
capabilities should make changes to ToS. Improper changes to
ToS can degrade network performance.
The IP trunk card transfers T.30 protocol (G3 Fax) implementations over
the IP network. Near real-time operational mode is supported where two
T.30 facsimile terminals are able to engage in a document transmission in
which the T.30 protocol is preserved.
The trunk uses the T.38 protocol on the connection between a pair of IP
Trunk 3.01 (and later) nodes.
The call acts in the same way as a gateway-to-gateway H.323 call. The call
is set up using the normal voice call process (that is, the normal voice call
codec negotiation process occurs and the corresponding codec payloadsize
and jitter buffer values are used). When the call setup is complete, the two
G3 Fax terminals are linked. The DSP detects the fax call setup tones and
switches to handle the fax call. For the remainder of the call, the parameters
administered for the fax call are used (for example, payload size).
Some implications of the fax call setup process are as follows:
•
•
a voice codec must be configured, even if only fax calls will be made
both ends of the call must be able to negotiate to a common voice codec
for the calls to be successful
All T.30 session establishment and capabilities negotiation are carried out
between the telephones through the IP trunk cards over the IP Trunk 3.01
(and later) network using the T.38 protocol. In terms of the internet fax
service roles, the IP trunk card acts as both the fax on-ramp gateway and
the fax off-ramp gateway, depending on the call direction.
The on-ramp gateway demodulates the T.30 transmission received from the
originating G3 Fax terminal. The T.30 facsimile control and image data is
transferred in an octet stream structure, using a Real Time Protocol (RTP)
payload, over User Datagram Protocol (UDP) transport mechanism.
Signaling specified by H.323 V.2 protocol is used for IP Trunk 3.01 (and
later) to IP Trunk 3.01 (and later) call setup.
Modules supporting facsimile transmission are responsible for the following:
•
fax speed detection and adjustment
•
protocol conversion from G3 Fax to RTP payload for fax data transfer
•
T.30 fax protocol support
•
T.38 fax-over-IP protocol
•
V.21 channel 2 binary signaling modulation and demodulation
•
High-level Data Link Control (HDLC) framing
•
V.27 term (2400/4800 bps) high speed data modulation and
demodulation
If two ends support T.30 protocol, theyare compatible only if external factors
(for instance, delay and signal quality) permit. Only IP Trunk 3.01 (and
later) node fax calls are supported.
IP Trunk 3.01 (and later) supports a maximum fax speed of 14.4 Kbps.
Remote Access
Remote Access is supported on IP Trunk 3.01 (and later). Remote Access
allows an TM 3.1 user with no IP Trunk 3.01 (and later) data, including
Nortel support personnel, to manage the IP trunk card remotely.
Management and support of the IP Trunk 3.01 (and later) network depend
on IP networking protocols including SNMP, FTP, and Telnet. The Nortel
Netgear RM356 modem router or equivalent should be installed on the
ELAN subnet in order to provide remote support access for IP Trunk 3.01
(and later) and other IP-enabled Nortel products.
Remote Access79
V.29 (7200/9600 bps) high speed data modulation and demodulation
V.17 (14390 bps) high speed data modulation
V.21 channel 2 detection
Multi-channel operation support
The Nortel Netgear RM356 modem router integrates the functions of a V.90
modem, a PPP remote access server, an IP router, and a 4-port 10BaseT
Ethernet hub, and provides a range of security features that may be
configured so as to comply with the customer’s data network security policy.
Do not install a modem router on the ELAN subnet without the explicit
approval of the customer’s IP network manager. The RM356 modem router
is not secure unless it is configured correctly according to the customer’s
network security policy and practices.
3.01 (and later) data (whether stored locally or on an TM 3.1 server). Once
connected, the remote user can perform any operation available to that PC.
Remote Access
Remote Access is supported on the MMCS IP Gateway. Remote Access
allows a MAT user, such as Nortel Networks support personnel, with no ITG
data to manage the ITG card remotely.
The IP Trunk 3.01 (and later) architecture isolates the TLAN network
interface from the system. However, the system does not have direct access
to per-call statistics on the voice quality of the call. These statistics are
important for the purpose of the following:
•
make sure the network is providing the contractual service level
•
solve help desk inquiries or refund "bad call" charges
•
identify network problems and track network performance
IP Trunk 3.01 (and later) uses a Remote Authentication Dial In User Service
(RADIUS) client to transmit these statistics from the IP trunk card to a
network device:
•
The IP trunk card sends a Start record when a call begins.
•
The IP trunk card sends an End record when the call is released.
•
The End record contains QoS information and the amount of data sent.
•
Both records contain the Called and Calling Party numbers for call
identification.
•
The TM 3.1 Call Accounting application does not correlate RADIUS per
call statistics with the Meridian 1/CS 1000M CDR.
A network "listener" receives Start and End messages and stores the data.
Applications can retrieve the stored data for processing and presentation
to the user.
A RADIUS client on the IP trunk card allows per-call statistics of the IP
network call to be sent from the cards to a network listener. The client is
based on RFC2139, which defines the accounting portion of the RADIUS
protocol. The IP trunk card uses the authentication algorithm based on
RFC1321.
Configuration
Use TM 3.1 to configure the following RADIUS parameters:
Data is configured at the IP Trunk 3.01 (and later) node level and is
distributed to all the IP trunk cards associated with the IP Trunk 3.01 (and
later) node.
Messaging
The RADIUS client sends two records to the network listener: one when
the call is answered and one at the end of the call. The messages are
sent by the Follower card which processes the voice call (not the DCHIP or
Leader if they are not handling the voice data). The RADIUS protocol uses
UDP for message exchange. The client sends a message to the listener
and waits for an acknowledgment. If no acknowledgment is received,
the client re-transmits the record using the standard exponential backoff
theme. The data is stored on the IP trunk card until an acknowledgment is
received. When an acknowledgment is received, the data is discarded. The
client stores a maximum of 100 records. This allows two Start and two
End records for each of the 24 or 32 ports (depending on whether it is an
ITG-Pentium 24-port trunk card or a Media Card 32-port trunk card).
Per-call statistics support using RADIUS Client 81
key for authenticating RADIUS records (the key is maintained between
the RADIUS client and the RADIUS server)
Start record
The Start record is sent when the call is answered. It contains the following
fields:
•
Calling party number
•Originating IP address and port
•
Called party number
•
Destination IP address and port (of the actual card handling the call, not
the remote Leader)
•
Call start time
•
Call duration (time from call initiation to call answer)
•
Codec used
•
Orig/Term call side indication
•
Snapshot of remote Gateway’s QoS at time of call connect
The calling and called numbers (with their corresponding IP addresses) are
just that, regardless of which end is doing the originating. So the Follower
card on the originating side generates a RADIUS record with its own IP
address as the originating IP address. The terminating Follower also
generates a RADIUS record with that far end’s IP address as the originating
IP address and its own IP address as the destination address.
If the call is not answered or is rejected, only an End record is generated.
End record
The End record is sent when the call is released. It contains the following
fields:
•
•Originating IP address and port
•
•
•
•
•
•
•
•
Calling party number
Called party number
Destination IP address and port (of the actual card handling the call, not
the remote Leader)
Call start time
Call duration (time from call answer to call release)
Codec used
Orig/Term call side indication
Number of bytes transferred (sent octets/packets)
Number of packets transferred (sent octets/packets)
SNMP MIB
MIB-2 support
•
Snapshot of latency seen at the end of the call
•
Packet loss
•
Snapshot of remote Gateway’s QoS at time of call release
The End record is also sent for calls which are not answered or are rejected.
These records do not include the packet loss, number of bytes transferred,
number of packets transferred and latency.
SNMP is the protocol used to communicate TM 3.1 IP Trunk 3.01 (and later)
alarms or events. Support for the SNMP Management Information Bases
(MIB) on the IP trunk card is composed of two parts: the standard MIB-2
and extensions for the IP trunk card.
The WindNet agent supports both SNMP-V1 and V2c protocols.
IP Trunk 3.01 (and later) SNMP agent
The SNMP agent supports the Operation, Administration, and Maintenance
(OA&M) of IP Trunk 3.01 (and later), using TM 3.1. It can configure the IP
trunk card through file transfer services. The agent supports the SNMP-V1
protocol.
The SNMP agent provides the following capabilities:
•
Retrieval of system wide variables, such as:
— card state
— number of DSPs on the card
— number of available voice channels
— IP addresses
SNMP MIB83
— software version
— number of IP Trunk 3.01 (and later) nodes in fallback (that is, PSTN
operation)
•
Control of D-channel state, such as:
— enable
— disable
— release
— establish
•
Retrieval of DSP information, such as:
— DSP firmware
— DSP self-test status
— card reset
•
SNMP configuration (that is, community names and trap subscription)
— alarm generation through SNMP traps
•
File transfer, including configuration files, software upgrade, dialing plan
files, bootp files, activity log, and call trace files
Codec refers to the voice coding and compression algorithm used by the
DSPs on the IP trunk card. The G.XXX series of codecs are standards
defined by the International Telecommunications Union (ITU). Different
codecs have different QoS and compression properties. The specific
codecs and the order in which they are to be used for codec negotiation is
configured in TM 3.1.
When configuring the IP Trunk 3.01 (and later) node in TM 3.1, select the
image containing the needed codecs, and the preferred codec negotiation
order. The final codec used is determined by the codec negotiation process
with the far end during call setup. Parameters can be configured for each
codec in an image.
IP Trunk 3.01 (and later) supports the following codecs:
•
•
•
•
G.711
G.729AB
G.729B
G.723.1
G.711
G.729AB
G.729B
The G.711 codec delivers "toll quality" audio at 64 kbit/s. This codec is
optimal for speech quality, as it has the smallest delay and is resilient
to channel errors. However, it uses the largest bandwidth. The G.711
codec is the default codec if the preferred codec of the originating node
is not available on the destination IP Trunk 3.01 (and later) node. Voice
Activity Detection/Silence Suppression is configurable through TM 3.1. An
ITG-Pentium 24-port trunk card supports 24 channels per card with G.711.
A Media Card 32-port trunk card supports 32 channels per card with G.711.
The G.729AB codec is the default preferred codec when adding a new IP
Trunk 3.01 (and later) node in TM 3.1. This codec provides near toll-quality
voice at a low delay. The G.729AB codec uses compression at 8 kbit/s (8:1
compression rate). Optional B Voice Activity Detection/Silence Suppression
is configurable through TM 3.1. An ITG-Pentium 24-port trunk card supports
24 channels per card with G.729AB. A Media Card 32-port trunk card
supports 32 channels per card with G.729AB.
The G.729B codec uses compression at 8 kbit/s (8:1 compression rate).
Optional B Voice Activity Detection/Silence Suppression is configurable
through TM 3.1. An ITG-Pentium 24-port trunk card supports only 16
channels per card with G.729B due to higher DSP resources required for
this codec. The Media Card 32-port trunk card does not support G.729B.
The G.723.1 codec provides the greatest compression. Voice Activity
Detection/Silence Suppression is configurable through TM 3.1. An
ITG-Pentium 24-port trunk card supports 24 channels per card with
G.723.1. A Media Card 32-port trunk card supports 32 channels per card
with G.723.1.
Three downloadable DSP profiles support the codecs shown in Table 9
"Codecs supported by IP Trunk 3.01 (and later)" (page 85).
Table 9
Codecs supported by IP Trunk 3.01 (and later)
Security passwords85
Profile 1
32 ms. Echo Cancel Tail
24 ports/card for ITG-P
24-port card
32 ports/card for SMC
32-port card
Profile 2
32 ms. Echo Cancel Tail
24 ports/card for ITG-P
24-port card
32 ports/card for SMC
32-port card
Fax
Profile 3
32 ms. Echo Cancel Tail
16 ports/card for ITG-P
24-port card
Not supported for SMC
32-port card
Each codec supports one of three sets of parameters: one for DSP, one
for fax, and one for codec.
WARNING
The Media Card 32-port trunk card does not support Profile 3.
Security passwords
When Telneting into the ELAN network interface or using the debug port,
a password must be entered when prompted. Two levels of passwords
are used to prevent unauthorized data access. Unauthorized data access
occurs when an unauthorized individual is able to view or modify confidential
data, such as employee lists, password lists, and electronic mail. This
information can be used to bypass Direct Inward System Access (DISA)
restrictions and avoid charges.
The following are the two levels of passwords for IP Trunk 3.01 (and later):
•
•
Administrator level
The Administrator level is the most basic level of password. It provides
unrestricted access to all IP Trunk administration options and to most of the
IP trunk card level administration options. It does not, however, allow any
type of low-level diagnostics to be performed.
Technical support level
The Technical support level is for use by Nortel personnel only. It allows
low-level message monitoring and factory testing.
"Estimate voice traffic calculations" (page 95)
"Calculate the number of IP Trunk 3.01 (and later) ports required" (page 99)
"Calculate number of IP trunk cards required" (page 101)
"Calculate Ethernet and WAN bandwidth usage" (page 112)
"IP Trunk 3.01 (and later) LAN installation and configuration" (page 154)
"Basic setup of the IP Trunk 3.01 (and later) system" (page 154)
"IP trunk card connections" (page 154)
"Configure a system with separate subnets for voice and management" (page
155)
"Subnet configurations" (page 155)
"Selecting public or private IP addresses" (page 157)
"Single subnet option for voice and management" (page 157)
"Multiple IP Trunk 3.01 (and later) nodes on the same ELAN and TLAN
segments" (page 158)
"General LAN considerations" (page 158)
"ELAN and TLAN network interface half- or full-duplex operation" (page 158)
"TLAN subnet design" (page 159)
"Configure the TLAN subnet IP router" (page 159)
"Setting up the ELAN subnet" (page 160)
"How to avoid system interruption" (page 160)
The Meridian Integrated IP Telephony Gateway (ITG) system performs the
following actions:
•
compresses PCM voice
•
demodulates Group 3 fax
•routes the packetized data over a private internet, or intranet
•
provides virtual analog ISDN Signalling Link (ISL) TIE trunks between
Meridian 1 ESN nodes
•
enables interworking with other Nortel VoIP products such as CS
1000M, and Business Communication Manager (BCM)
IP Trunk 3.01 (and later) routes voice traffic over existing private IP network
facilities with available under-used bandwidth on the private WAN backbone.
IP Trunk 3.01 (and later) is targeted towards the Enterprise customer who
has a Meridian 1/CS 1000M system installed for providing corporate voice
services and an intranet for corporate data services. A customer is expected
to use the IP Trunk 3.01 system to move traffic from a PSTN-based network
to the intranet. Voice and fax services which depended on circuit-switched
and Time Division Multiplexing (TDM) technology are transported using
packet-switched and statistical multiplexing technology.
This chapter provides guidelines for designing a network of IP Trunk 3.01
(and later) nodes over the corporate intranet. It describes how to qualify the
corporate intranet to support an IP Trunk 3.01 (and later) network and how
to determine changes required to maintain the quality of voice services
when moving those services from the PSTN. It addresses requirements for
the successful integration with the customer’s existing LAN. By following
these guidelines, the IP Trunk 3.01 (and later) network can be designed
so that the cost and quality tradeoff is at best imperceptible and at worst
within a calculated tolerance.
Pre-installation analysis of the data network enables IP Trunk 3.0 (and later)
to be provisioned correctly. For proper analysis and deployment, obtain a
network diagram or a description of the network topology and hierarchy.
Nortel recommends using a data network analyzer (for example, Sniffer
for evaluation and troubleshooting.
TM
)
Audience
This chapter is intended for telecom and datacom engineers who design and
install the IP Trunk 3.01 (and later) node portion of the VoIP network. It is
assumed that the telecom engineer is familiar with engineering the Meridian
1/CS 1000 system and obtaining system voice and fax traffic statistics. It is
assumed that the datacom engineer is familiar with the intranet architecture,
LAN installations, tools for collecting and analyzing data network statistics,
and data network management systems.
For information on designing a Meridian 1/ CS 1000 network, refer to the
following NTP:
•
•
•
•
Communication Server 1000M and Meridian 1 Small System Planning
and Engineering (NN43011-220)
Communication Server 1000M and Meridian 1 Large System Planning
and Engineering (NN43021-220)
Communication Server 1000S: Planning and Engineering
(NN43031-220)
CommunicationServer 1000E Planning and Engineering (NN43041-220)
The IP Trunk 3.01 (and later) system was designed for operation on a
well-provisioned, stable LAN. Delay, delay variation or jitter, and packet loss
must be minimized end-to-end across the LAN and WAN. The design and
configuration of the LAN and WAN that link the IP Trunk 3.01 (and later)
system must be determined. If the intranet becomes overloaded, new calls
to the IP Trunk 3.01 (and later) system fall back to normal circuit-switched
voice facilities so that the Quality of Service (QoS) does not degrade for
new calls.
IP Trunk 3.01 (and later) is for intranet use only. IP Trunk 3.01 (and later)
provides virtual analog ISL TIE trunks between two Meridian 1 systems
in an ESN network, as shown in Figure 23 "The IP Trunk 3.01 (and later)
intranet" (page 91). IP Trunk 3.01 (and later) does not support modem
traffic except for Group 3 fax. The technician must configure the Meridian
1/ CS 1000M routing controls to route modem traffic over circuit-switched
trunks instead of over IP Trunk 3.01 (and later).
Figure 23
The IP Trunk 3.01 (and later) intranet
Introduction 91
IP Trunk 3.01 (and later) is available for the following systems running CS
1000 Release 4.0 software:
•
Meridian 1 PBX 61C CP PII
•
Meridian 1 PBX 81C CP PII
•
Meridian 1 PBX 11C Chassis
•
CS 1000M SG
•
CS 1000M MG
•
CS 1000M Cabinet
•
CS 1000M Chassis
The IPE trunk cards plug into the Meridian 1/CS 1000M IPE shelf.
A maximum of eight ITG-Pentium 24-port trunk cards can fit on one IPE
shelf. Each card takes up two slots on the IPE shelf.
A maximum of 16 Media Card 32-port trunk cards can fit on one IPE
shelf. Each IP trunk card takes up one slot on the IPE shelf. For Class B
compliance to EMC regulations, only 10 Media Card 32-port trunk cards can
be placed on an IPE shelf. For Class A compliance, there are no limitations
on the Media Card 32-port trunk card. For more information, see Appendix
"Environmental and electrical regulatory data" (page 449).
An IPE shelf can contain a mixture of ITG-Pentium 24-port trunk cards
and Media Card 32-port trunk cards.
Cabinet systems operating under Class B Electro-Magnetic Compatibility
(EMC) standards can only hold a total of two IP Trunk cards, divided
between the main and expansion cabinets. This can be extended to two
cards in each main or expansion cabinet if all cabinets are separated from
each other by at least ten meters distance. For Cabinet systems operating
under Class A EMC standards, there are no restrictions.
Scope
For Meridian 1 Option 11C Cabinet, Meridian 1 PBX 11C Chassis,
CS 1000M Cabinet, and CS 1000M Chassis systems, the SDI/DCH
(NTAK02BB) card occupies one slot on the cabinet and is connected to
the IP trunk card through the backplane. Only ports 1 and 3 are available
for use as DCHI.
The IP trunk card uses a 10BaseT Ethernet network interface located on
the card backplane I/O connector to carry IP Trunk 3.01 (and later) system
management traffic; it connects to the ELAN subnet.
These engineering guidelines address the design of the IP Trunk 3.01 (and
later) network, which consists of the following:
•
IP Trunk 3.01 (and later) nodes
•
Telephony LAN (TLAN) subnets to which the IP Trunk 3.01 (and later)
nodes are connected
•A corporate intranet which interconnects the various TLAN subnets
These guidelines require that the customer has a corporate intranet in
place that spans the sites where the IP Trunk 3.01 (and later) nodes are to
be installed.
Previously, Meridian 1 networks depended on voice services such as
LEC and IXC private lines. With IP Trunk 3.01 (and later) technology, the
Meridian 1 and CS 1000 systems can select a new delivery mechanism, one
that uses packet-switching over a data network or corporate intranet. The
role of the IP Trunk 3.01 (and later) node is to convert steady-stream digital
voice into fixed-length IP packets, provide ISDN signalling, and translate
PSTN numbers into IP addresses. The IP packets are transported across
the IP data network with a low latency that varies with strict limits.
The term "voice services" also includes fax services.
IP evolved from a protocol that allowed multi-vendor hosts to communicate.
The protocol adopted packet-switching technology, providing bandwidth
efficiency for bursty data traffic that can tolerate high latency and jitter
(variation in latency). Since IP supported the TCP transport layer, which
provided connection-oriented and reliable transport, IP took on the
properties of being connectionless and a best-effort delivery mechanism.
The TCP/IP paradigm worked well in supporting data applications at that
time.
New considerations come into play now when the same corporate network
is expected to deliver voice traffic. The intranet introduces impairments,
delay, delay variation, and data packet loss, at levels that are higher than
those delivered by voice networks. Delay between talker and listener
changes the dynamics and reduces the efficiency of conversations, while
delay variation and packet errors causes introduces glitches in conversation.
Connecting the IP Trunk 3.01 (and later) nodes to the corporate intranet
without preliminary assessments and QoS mechanisms can result in
unacceptable degradation in voice service. Correct design procedures and
principles must be considered.
A good design for the IP Trunk 3.01 (and later) network must begin with
an understanding of traffic and the underlying network that will transmit
the traffic. See Figure 24 "IP Trunk 3.01 (and later) network engineering
Figure 24
IP Trunk 3.01 (and later) network engineering process
Three preliminary steps must be undertaken.
1. Calculate IP Trunk 3.01 (and later) traffic. Estimate the amount of
traffic that the system will route through the IP Trunk 3.01 (and later)
network. This total must include the estimated traffic between the IP
trunk cards and the Signaling Server. This in turn places a traffic load
on the corporate intranet. This is described in "IP Trunk 3.01 (and later)
2. Assess WAN link resources. If resources in the corporate intranet are
not sufficient to adequately support voice services, the cause is usually
insufficient WAN resources. "Assess WAN link resources" (page 122)
outlines how this assessment can be made.
3. Measure the existing intranet’s Quality of Service (QoS). Estimate the
quality of voice service the corporate intranet can deliver. "Measure
intranet QoS" (page 140) describes how to measure prevailing delay
and error characteristics of an intranet.
After the assessment phase, the IP Trunk 3.01 (and later) network can be
designed and implemented. This design not only involves the IP Trunk
3.01 (and later) elements, but can also require making design changes to
the existing customer intranet. "Fine-tune network QoS" (page 133) and
"Implement QoS in IP networks" (page 127) provide guidelines for making
modifications to the intranet.
IP Trunk 3.01 (and later) traffic engineering
To design a network is to size the network so that it can accept a calculated
amount of traffic. The purpose of the IP Trunk 3.01 (and later) network is
to deliver voice traffic that meets QoS objectives. Since traffic determines
network design, the design process must start with obtaining an offered IP
Trunk 3.01 (and later) traffic forecast. The traffic forecast drives drive the
following:
•
IP Trunk 3.01 (and later) hardware requirements
•
WAN requirements
•
TLAN subnet requirements
Traffic forecasting is a process that often requires several tries to achieve
satisfactory results. For example, a WAN might not have enough bandwidth
to support all the IP trunks required; therefore the codec choice or the
number of trunks provisioned must be adjusted.
Estimate voice traffic calculations
Follow the steps in Procedure 4 "Estimating voice traffic" (page 95) to
calculate an estimate of voice traffic.
Procedure 4
Estimating voice traffic
StepAction
1
Calculate Voice on IP traffic.
CCS/user=# of calls/ * Average Holding Time (in seconds)/100
Total voice CCS (Tv) = CCS/user x No. of VoIP users
The number of VoIP users (telephones) is the potential population
in the system that can generate/receive traffic through the IP Trunk
3.01 (and later) node. This number may be estimated for a new
Meridian 1 customer.
If the installation is for an existing customer, base the VoIP traffic on
measured route traffic from traffic report TFC002, which provides
CCS for each route. A customer must determine the amount of
expected private network voice traffic.
2
Calculate Fax on IP traffic
CCS/user sending fax = # of pages sent/fax * Average Time to send
a page (default 48 seconds)/100
CCS/user receiving fax = # of pages received/fax * Average Time
to receive a page (default 48 seconds)/100
Total fax CCS (Tx) = CCS/fax sent*No. of users sending fax +
CCS/fax received* No. of users receiving fax
The user sending or receiving a fax can be the same person or
different persons. It is the number of faxed documents and the
average number of pages per faxed document that are important.
The time unit for fax traffic is also the busy hour. The busy hour
selected must be the hour that gives the highest combined voice
and fax traffic.
3
Total the ITG CCS.
Total IP Trunk 3.01 (and later) traffic (T) = Tv + Tx
4
Refer to Poisson P.01 table to find IP Trunk 3.01 (and later) ports
required to provide a blocking Grade of Service of 1% assuming
Poisson random distribution of call origination and zero correlation
among calls.
A lower Grade of Service, such as P.10, may be preferred if overflow
routing is available through the PSTN, circuit-switched VPN, or ITG
ISL TIE trunks.
For P.01 blocking Grade of Service, the number of trunks (IP Trunk
and Tx/36 indicate the average number of simultaneous callers.
Nortel Communication Server 1000
IP Trunk Fundamentals
NN43001-563 01.01 Standard
Release 5.0 30 May 2007
Page 97
IP Trunk 3.01 (and later) traffic engineering97
This calculation requires perfectly queued and perfectly smooth
traffic.
Tv/36*bandwidth output per port = voice bandwidth per node (Bv)
Tx/36*bandwidth output per port = fax bandwidth per node (Bx)
Total bandwidth (Bt) = Bv + Bx
For WAN calculation, consider only the larger of fax traffic sent or
received.
6
Adjust requirement for traffic peaking.
Peak hour bandwidth per node = Bt*1.3 (default)
—End—
Procedure 5 "Calculating IP Trunk 3.01 (and later) port and bandwidth
requirements" (page 97) is used to calculate IP Trunk 3.01 (and later)
port and, therefore, IP network bandwidth requirements. In the WAN
environment, the traffic parcel is defined for each destination pair (route).
The total node traffic should be sub-divided into destination pair traffic. The
rest of the calculation procedure continues to apply.
Example 1:
IP Trunk 3.01 (and later) ports and bandwidth engineering (Silence
Suppression enabled)
In this configuration example of 120 VoIP users, each user generates four
calls using the IP network (originating and terminating) with an average
holding time of 150 seconds in the busy hour.
In the same hour, 25 faxes were sent and 20 faxes received. The faxes sent
averaged 3 pages, while the faxes received averaged 5 pages. The average
time to set up and complete a fax page delivery is 48 seconds.
The codec of choice is G.729AB, voice packet payload is 30 ms. The fax
modem speed is 14.4 kbit/s and payload is 16.6 ms. How many IP Trunk
3.01 (and later) ports are needed to meet P.01 blocking Grade of Service?
What is the traffic in kbit/s generated by this node to the TLAN subnet?
Follow the steps in Procedure 5 "Calculating IP Trunk 3.01 (and later) port
and bandwidth requirements" (page 97) to calculate IP Trunk 3.01 (and
later) port and bandwidth requirements.
Procedure 5
Calculating IP Trunk 3.01 (and later) port and bandwidth requirements
CCS/fax sent = 3*48/100 = 1.44 CCS
CCS/fax received = 5*48/100 = 2.4 CCS
Total fax CCS (Tx + Rx) = 1.44*25 + 2.4*20 = 36+ 48= 84 CCS
3Calculate IP Trunk 3.01 (and later) traffic during busy hour.
Total traffic (T) = Tv + Tx = 720 + 84 = 804 CCS
4
Refer to the Poisson P.01 table (Table 10 "Trunk traffic – Poisson 1%
blocking Grade of Service" (page 99)) to find the number of IP Trunk
3.01 (and later) ports required for 1% blocking Grade of Service.
For P.10 blocking Grade of Service, refer to Table 11 "Trunk traffic –
Poisson 10% blocking Grade of Service" (page 100).
804 CCS can be served by 35 IP Trunk 3.01 (and later) ports
with P.01 blocking Grade of Service. Two ITG-Pentium 24-port
trunk cards are needed to serve this customer.
5
Calculate average bandwidth use on the TLAN subnet.
For voice:
720/36*30.7 = 614 kbit/s
For fax:
84/36*46.1 =108 kbit/s
Total bandwidth = 614 + 108 = 722 kbit/s
6
This example is based on the G.729AB codec with 30 ms payload size and
Silence Suppression enabled. For relations of user-selectable parameters
such as payload size, codec type, packet size and QoS, refer to "Set QoS
Peak hour bandwidth requirement = 722*1.3 = 939 kbit/s
This is the spare bandwidth a TLAN subnet requires to transmit
the VoIP and fax traffic. Nortel recommends that the TLAN
subnet handle IP Trunk 3.01 (and later) traffic exclusively.
—End—
Nortel Communication Server 1000
IP Trunk Fundamentals
NN43001-563 01.01 Standard
Release 5.0 30 May 2007
Page 99
IP Trunk 3.01 (and later) traffic engineering99
Calculate the number of IP Trunk 3.01 (and later) ports required
IP Trunk 3.01 (and later) TIE trunks are provisioned based on average
busy-hour traffic tables, using the calculated amount of voice and fax
traffic between IP Trunk 3.01 (and later) nodes. Table 10 "Trunk traffic
– Poisson 1% blocking Grade of Service" (page 99) shows the number
of trunks required based on average busy hour CCS for a 1% blocking
Grade of Service. Table 11 "Trunk traffic – Poisson 10% blocking Grade of
Service" (page 100) shows the number of trunks required based on average
busy-hour CCS for a 10% blocking Grade of Service.
A lower Grade of Service, such as P.10, might be preferred if overflow
routing is available through the PSTN, circuit-switched VPN, or IP Trunk
3.01 (and later) TIE trunks.
Table 10
Trunk traffic, Poisson 1 per cent blocking Grade of Service