Routing Mechanisms

This section describes several call routing mechanisms.

Username/SIP URI Routing

Username/SIP URI routing allows routing of requests based on the username and/or domain name in the SIP Request-URI. The username may be digits or an actual username. In either case, routing tries to find the most specific match for the (full) username and domain name, followed by less specific matches on suffixes of the domain name.

By default, the SBC Core applies the default time range profile “ALL” to a username route, which specifies all days and times.

Standard Destination-Based Routing

The Standard destination based routing supports route lookups based on following parameters:

• Carrier based routing
• Trunk group based routing
• Calling number based routing
• Called number based routing

Enhanced 911 Emergency Routing

In a VoIP environment, subscribers can take their phone numbers outside of their original geographic location (local number portability) or use a VoIP adapter to make calls from a remote location. To make an emergency services available to subscribers not tied to fixed geographic locations, Enhanced 911 tracks caller location and the Public Services Answering Point nearest to them by means of a VoIP Positioning Center (VPC).

The SBC supports Enhanced 911 emergency services using both IPv4 and IPv6 IP addresses when sending Emergency Services Routing requests to an external VPC. The VPC reply contains the routing information for the appropriate emergency services dispatcher nearest the VoIP subscriber dialing 911. SIP to SIP-I call routing is also supported.

Leading Digit Routing

SBC standard routing performs a leading digits prefix match to determine the best route available. Routes are defined based on either the entire called number or the leading digits of the number.

Route Simulation

The SBC supports the capability to configure test routes for routes simulation using the ‘routingLabelRoute’ parameter and ‘testing’ sub-parameter of global call routing feature.

CLI command example:

% set global callRouting routingLabel 1 routingLabelRoute 2 testing test

Where testing parameter has three options:

• nonTest – When selected, the ERE does not return the route when the CPC value in a policy request is Test Call. When the CPC value is not a Test Call or is not present in the policy request, the ERE returns the route. Select this option to identify routes not to be returned during testing.
• normal – After testing and verifying a route, select this option to use the route for live calls. When selected, the ERE returns the route regardless of the calling party category (CPC) value, or absence of a CPC value, received in the policy request. (Default setting).
• test – When the CPC value in a policy request is Test Call, the ERE returns the route. When the CPC value is not Test Call or is not present in the policy request, the ERE does not return the route.

Route Prioritization

The SBC supports route prioritization in a routing label based on the methods:

• Sequence—The SBC allocates the routes in the order of the values provided in the route Sequence field.
• Proportion—The SBC uses the Proportion method to determine the first route on a call-by-call basis.

For example, if the Route Label contains three routes with assigned proportions of 80, 50, and 30, the system generates a random number (in this case between 0 and 159), and if the number is between 0 and 79 the system sets the route with the 80 proportion as the first route. If the random number is between 80 and 129, the system sets the route with the 50 proportion as the first route. If the random number is between 130 and 159, the system sets the route with the 30 proportion as the first route.

• Round Robin—The SBC distributes call traffic equally across the routes in a Routing Label. For each call, the routes are cyclically rotated by one position.

For example, call 1 receives routes 1-10, call 2 receives routes 2-11, call 3 receives routes 3-12, etc.

• Route cost—The SBC determines the routes by cost, the least (lowest) cost route being the first priority route selected. In the case of equal cost routes, use the Route Prioritization Type for Equal Cost Routes parameter to select a secondary route prioritization type.

HD Codec Based Routing

The HD codec based prioritization method is invoked after executing the existing route prioritization methods.

Prerequisites

HD codec based prioritization is an end-to-end routing feature on the ingress offer. To configure the codec priority to remain constant throughout SBC processing (e.g., wide band remains the highest priority codec), enable the following settings in Packet Service Profile screen.

• Honor Remote Precedence—Select this option in the ingress route PSP so that the priority specified by the ingress peer in the offer is unchanged when sent to the egress.
• Send Route PSP Precedence—Clear this option in the ingress route PSP so that the priority specified by the ingress peer in the offer is unchanged when sent to the egress.

Prioritizing routes based on HD codec

In Codec based prioritization, the PSX prioritizes the egress routes for the following High Definition (HD) codecs:

• G.722
• G.722.1
• G.722.2 (AMR-WB)
• G729.1
• SILK
• MSRT
• EVS

The PSX receives the list of HD codecs and their sampling rates in the policy request sent by the SBC. If the ingress PSP contains any of the above mentioned HD codecs, then the PSX reorders the egress routes as follows:

• Routes with first priority egress codecs are placed on top
• Routes with the second priority egress codecs are placed next
• All the remaining routes are placed at the bottom

If the ingress PSP does not contain any of the above mentioned codecs, then the PSX configures the routes normally based on the existing logic.

TGRP/Trunk-Context Based Routing

A SIP-to-PSTN gateway can have trunks connected to different carriers. Plus, a SIP proxy may choose (based on proprietary routing logic) a carrier in which a call is sent when it proxies a session setup request to the gateway. Since, multiple carriers can transport a call to a particular phone number, a phone number by itself is not sufficient to identify the carrier at the gateway. To overcome this, the ERE routing logic uses “tgrp” and “trunk-context” parameters in the Request URI header as described below.

The SBC supports processing Destination Trunk Group (DTG) and Originating Trunk Group (OTG) parameters received in Contact and Request-URI messages, and transparently passing this information or inserting it in response messages or egress INVITE headers. When configured this way, even if OTG or DTG values are included in the received messages they are not used in processing of trunk group values for response and egress.

If the PES receives destination trunk group parameters (tgrp and trunk-context) in the policy request and if “Process Destination Tgrp” and “Process Destination Trunk Context” flags are enabled for the ingress trunk group, PES performs a light policy lookup and skips full policy dip. In this scenario, the ERE uses the trunk group name received in the tgrp parameter (similar to the dtg parameter). Also, the trunk-context parameter value is ignored. If DTG is present, it is also ignored.

The IP address is unique within a zone. To perform a reverse lookup against an IP peer, the IP address and zone are required. When trunk-context contains an IP address, use a default zone to look up the IP peer.

Two options are available to identify the zone name:

1. Use the same zone on which incoming call is received.
2. Use address context to look up the IP peer based on assumption that ingress peer and IP address in trunk-context belong to the same address context. The following details are sent in the policy response:

• • The name of the SBC that is processing the request
  • Zone name
  • IP peers IP Address/ FQDN

Since the local trunk group is not returned in the policy response, the local trunk group in the zone is determined by the SBC to ensure the call is sent to the destination IP Peer. The Trunk Resource Manager (TRM) looks up the Network Selector Table (NST) associated with the zone to determine the local trunk group.

After determining IP peer based on the trunk-context, the zone name to which this IP peer belongs is determined by fetching IP peer table. The ERE can send back the egress zone information in the policy response by populating zone ID in the zoneIndex attribute in the route4Attributes table.

A CLI configuration example is shown below.

• Allow processing of “tgrp” and “trunk-context” for feature control profile “DEFAULT_IP”:

  % set profiles featureControlProfile DEFAULT_IP processDestinationTrunkGroupAndTrunkContext enable

• Associate the profile with a SIP trunk group:

  % set addressContext default zone defaultSigZone sipTrunkGroup DEFAULT_TG policy featureControlProfile DEFAULT_IP

tgrp and trunk-context Parameter Taken into Account Together

Trunk Groups are identified by two parameters: tgrp and trunk-context.

The trunk-context establishes the scope of the trunk group specified in the tgrp parameter and determines if it is for a single gateway, a set of gateways, or an entire domain managed by the service provider.

The value in the trunk-context identifies a gateway.

When the remote party sends the invite with both tgrp = egress_tg, trunk_context = Egress IP / Egress IP + Port / FQDN, the SBC chooses the egress trunk group based on the tgrp parameter in the Request-URI (R-URI). The SBC uses the trunk-context as destination, if there is no IP Peer or IP Signaling Peer Group is not configured.

 The SBC does not provision the IP peer when the remote IP address is already provisioned in some other system and the user does not want duplicate provisioning, or when the user does not know the remote IP address.

 You can use a trunk group for multiple destinations without having the IP address in its list of ingress prefixes.

How it works

When you disable processTgrpContext at the SBC, the SBC sends tgrp and trunk-context to the PSX.

In this case, when you enable the processDestinationTgrp in Feature Control Profile, the PSX does light policy lookup and skips all the SSG-based services, number translation, and routing functions.

If you configure the ipSignalingPeerGroup in the Trunk Group, the SBC routes the request to the configured ipSignalingPeerGroup.

If you do not configure the ipSignalingPeerGroup/ipPeer in the Trunk Group and the trunk-context is available, then the SBC routes the request to trunk-context.

Process tgrp/trunk-context in R-URI On Light Dip

Prior to SBC 8.0, the SBC supported the tgrp/trunk-context functionality in R-URI for INVITE only. The SBC now supports the tgrp/trunk-context functionality in R-URI for NON-INVITE messages (same as DTG).

The light dip principle instructs the Diameter+ PSX to skip most of the services and return a single route for the specified trunk group. This is useful for networks where an upstream SIP device  (for example, the PSX Proxy which anchors the SIP session) performs crank back / route advance and sends the profile.

For more information, refer to the following PSX documents:

• Dynamic Routing and Policy Selection Based on D+
• OTG and DTG Parameter Processing
• Processing Trunk Group and Trunk-Context Parameters in SIP Messages  
• Provisioning Trunk Context and R-URI in Trunk Group  
• PSX as Proxy Server

The SBC uses the R-URI tgrp/trunk-context as a replacement for the Destination Trunk Group (DTG) parameter. With this enhancement, the SBC treats the tgrp parameter for INVITE and NON-INVITE transactions ( the same as if it received the DTG parameter) allowing a light dip at the D+ PSX, for all INVITE and NON-INVITE out-of-dialog SIP transactions.

The following figure depicts the example of NON-INVITE OOD transaction.

Figure 1: NON-INVITE OOD transaction

Routing Calls to SIP Server Using defaultSigZone

The SBC is capable of routing calls to an SIP Server using defaultSigZone, a named IP trunk group and a sipSigPort in the defaultSigZone.

When the SBC queries the PSX for routing a call to the SIP Server (no zone or Zone Index 0 configured on the PSX), the SIP Server IP address, virtual trunk group and Zone Index 0 is returned in the response. The SBC will map PSX Zone Index 0 to the SBC defaultSigZone, and use customer-provisioned named sipTrunkGroup (the ingress IP Prefix of which matches the SIP Server IP address) and customer-provisioned sipSigPort in the defaultSigZone for routing the call.

Figure 2: SBC - SIP Server Network Diagram

Example:

SIP Trunk Group

% set addressContext default zone defaultSigZone sipTrunkGroup defaultSipTrunkGroup ingressIpPrefix 10.7.6.40 32
% set addressContext default zone defaultSigZone sipTrunkGroup defaultSipTrunkGroup media mediaIpInterfaceGroupName DLIG
% set addressContext default zone defaultSigZone sipTrunkGroup defaultSipTrunkGroup state enabled mode inService

SIP Signaling Port

% set addressContext default zone defaultSigZone sipSigPort 1 ipAddressV4 10.7.15.146 portNumber 5060 transportProtocolsAllowed sip-udp
% set addressContext default zone defaultSigZone sipSigPort 1 ipInterfaceGroupName DLIG
% set addressContext default zone defaultSigZone sipSigPort 1 mode inService state enabled

Refer to Zone - SIP Trunk Group - CLI for command syntax and parameter descriptions.

Multiple Named Non-Routable IPTGs in defaultSigZone

The SBC Core also supports multiple named non-routable IP trunk groups created in the defaultSigZone for GW-GW, SIP Servers and H.323 calls. The following routes are supported in defaultSigZone and default addressContext:

GW-GW Routing

Table 1: GW-GW Routing

Routing to SIP Servers

(SIP hop-by-hop routing and SIP core optimized routing)

Table 2: Routing to SIP Servers

Routing to H.323 Servers

Table 3: Routing to H.323 Servers

 Figure 3: Using Multiple Named Non-Routable IPTGs in defaultSigZone for Calls