VoIP Configuration

The important thing to remember is that QoS must be configured throughout your network---not just on your router running VoIP---to improve voice network performance. Not all QoS techniques are appropriate for all network routers. Edge routers and backbone routers in your network do not necessarily perform the same operations; the QoS tasks they perform might differ as well. To configure your IP network for real-time voice traffic, you need to consider the functions of both edge and backbone routers in your network and then select the appropriate QoS tool or tools.

• Packet classification

• Admission control

• Bandwidth management

• Queuing

• High-speed switching and transport

• Congestion management

• Queue management

Scalable QoS solutions require cooperative edge and backbone functions.

RSVP works well on PPP, HDLC, and similar serial line interfaces. It does not work well on multi-access LANs. RSVP can be equated to a dynamic access list for packet flows.

You should configure RSVP to ensure QoS if the following conditions describe your network:

• Small scale voice network implementation

• Links slower than 2 Mbps

• Links with high utilization

• Need for the best possible voice quality

This command starts RSVP and sets the bandwidth and single-flow limits. The default maximum bandwidth is up to 75 percent of the bandwidth available on the interface. By default, the amount reservable by a flow can be up to the entire reservable bandwidth.

On subinterfaces, RSVP applies to the more restrictive of the available bandwidths of the physical interface and the subinterface.

Reservations on individual circuits that do not exceed the single flow limit normally succeed. However, if reservations have been made on other circuits adding up to the line speed, and a reservation is made on a subinterface that itself has enough remaining bandwidth, it will still be refused because the physical interface lacks supporting bandwidth.

A Cisco 1750 running VoIP and configured for RSVP requests allocations using the following formula:

bps=packet_size+ip/udp/rtp header size * 50 per second
 

For G.729, the allocation works out to be 24,000 bps. For G.711, the allocation is 80,000 bps.

For more information about configuring RSVP, refer to the "Configuring RSVP" chapter of the Network Protocols Configuration Guide, Part 1 for Cisco IOS Release 12.0T.

In general, multilink PPP with interleaving is used in conjunction with weighted fair queuing and RSVP or IP precedence to ensure voice packet delivery. Use multilink PPP with interleaving and weighted fair queuing to define how data is managed; use RSVP or IP precedence to give priority to voice packets.

You should configure multilink PPP if the following conditions describe your network:

• Point-to-point connection using PPP encapsulation

• Links slower than 2 Mbps

Note Do not use multilink PPP on links greater than 2 Mbps.

Multilink PPP support for interleaving can be configured on virtual templates, dialer interfaces, and ISDN BRI or PRI interfaces. To configure interleaving, you need to complete the following tasks:

• Configure the dialer interface or virtual template, as defined in the relevant chapters of the Dial Solutions Configuration Guide for Cisco IOS Release 12.0T.

• Configure multilink PPP and interleaving on the interface or template.

To configure multilink PPP and interleaving on a configured and operational interface or virtual interface template, use the following interface configuration commands:

Step	Command	Task
1.	ppp multilink	Enable Multilink PPP.
2.	ppp multilink interleave	Enable real-time packet interleaving.
3.	ppp multilink fragment-delay milliseconds	Optionally, configure a maximum fragment delay of 20 milliseconds.
4.	ip rtp reserve lowest-UDP-port range-of-ports [maximum-bandwidth]	Reserve a special queue for real-time packet flows to specified destination UDP ports, allowing real-time traffic to have higher priority than other flows. This only applies if you have not configured RSVP.

Note You can use the ip rtp reserve command instead of configuring RSVP. If you configure RSVP, this command is not required.

For more information about multilink PPP, refer to the "Configuring Media-Independent PPP and Multilink PPP" chapter in the Dial Solutions Configuration Guide for Cisco IOS Release 12.0T.

You should configure RTP header compression if the following conditions describe your network:

• Links slower than 2 Mbps

• Need to save bandwidth

Note Do not use RTP header compression on links greater than 2 Mbps.

Perform the following tasks to configure RTP header compression for VoIP. The first task is required; the second task is optional.

16384 + (4 x number of voice ports in the router)
 

Custom Queuing and other methods for identifying high priority streams should be configured for these port ranges. For more information about custom queuing, refer to the "Managing System Performance" chapter in the Configuration Fundamentals Configuration Guide for Cisco IOS Release 12.0T.

In general, weighted fair queuing is used in conjunction with multilink PPP with interleaving and RSVP or IP precedence to ensure voice packet delivery. Use weighted fair queuing with multilink PPP to define how data is managed; use RSVP or IP precedence to give priority to voice packets. For more information about weighted fair queuing, refer to the "Managing System Performance" chapter in the Configuration Fundamentals Configuration Guide for Cisco IOS Release 12.0T.

Note You can use a period (.) to represent variables (such as extension numbers) in a telephone number. A period is similar to a wildcard, which matches any entered digit.

The information included in this example needs to be configured on both Cisco 1750 Router 1 and Cisco 1750 Router 2. In this configuration, Cisco 1750 Router 1 can call any number string that begins with the digits 17295552 or 17295553 to connect to Cisco 1750 Router 2. Similarly, Cisco 1750 Router 2 can call any number string that begins with the digits 14085551 to connect to Cisco 1750 Router 1.

For inbound call legs, a dial peer might be associated with the calling number or the voice-port number. Outbound call legs always have a dial peer associated with them. The destination pattern is used to identify the outbound dial peer. The call is associated with the outbound dial peer at setup time.

POTS dial peer associate a telephone number with a particular voice port so that incoming calls for that telephone number can be received and outgoing calls can be placed. VoIP dial peers point to specific devices (by associating destination telephone numbers with a specific IP address) so that incoming calls can be received and outgoing calls can be placed. Both POTS and VoIP dial peers are needed to establish VoIP connections.

To enter the dial peer configuration mode (and select POTS as the method of voice-related encapsulation), use the following global configuration command:

Router(config)# dial-peer voice number pots
 

The number value of the dial-peer voice pots command is a tag that uniquely identifies the dial peer. (This number has local significance only.)

To configure the identified POTS dial peer, use the following dial peer configuration command:

Router(config-dial-peer)# destination-pattern string
 

The string value of the destination-pattern command is the destination telephone number associated with this POTS dial peer.

To enter the dial peer configuration mode (and select VoIP as the method of voice-related encapsulation), use the following global configuration command:

Router(config)# dial-peer voice number voip
 

The number value of the dial-peer voice voip command is a tag that uniquely identifies the dial peer.

To configure the identified VoIP dial peer, use the following dial peer configuration commands

:

	Command	Task
Step 1	destination-pattern string	Define the destination telephone number associated with this VoIP dial peer.
Step 2	session target {ipv4:destination-address | dns:host-name}	Specify a destination IP address for this dial peer.

• If you have relatively few dial peers configured, you can use the show dial-peer voice command to verify that the data configured is correct. Use this command to display a specific dial peer or to display all configured dial peers.

• Use the show dialplan number command to show which dial peer is reached when a particular number is dialed.

• Ping the associated IP address to confirm connectivity. If you cannot successfully ping your destination, refer to the "Configuring IP" chapter in the Network Protocols Configuration Guide, Part 1 for Cisco IOS Release 12.0T.

• Use the show dial-peer voice command to verify that the operational status of the dial peer is up.

• Use the show dialplan number command on the local and remote routers to verify that the data is configured correctly on both.

• If you have configured number expansion, use the show num-exp command to check that the partial number on the local router maps to the correct full E.164 telephone number on the remote router.

• If you have configured a CODEC value, there can be a problem if the VoIP dial peers on either side of the connection have incompatible CODEC values. Make sure that both VoIP peers have been configured with the same CODEC value.

Caution If you are not familiar with Cisco IOS debug commands, you should read the "Using Debug Commands" section in the "VoIP Debug Commands" chapter before attempting any debugging.

Each VIC is specific to a particular signaling type; therefore, VICs determine the type of signaling for the voice ports. Voice-port commands define the characteristics associated with a particular voice-port signaling type.

• Pick up the handset of an attached telephony device and listen for a dial tone.

• Check for DTMF detection if you have a dial tone. If the dial tone stops when you dial a digit, the voice port is configured properly.

• Use the show voice-port command to verify that the data configured is correct.

• Ping the associated IP address to confirm connectivity. If you cannot ping your destination, refer to the Network Protocols Configuration Guide, Part 1 for Cisco IOS Release 12.0T.

• Use the show voice-port command to make sure that the port is enabled. If the port is offline, use the no shutdown command.

• Make sure the VICs are correctly installed. For more information about installing a VIC in your router, refer to the <Emphasis>Cisco WAN Interface Cards Hardware Installation Guide.

If you need to change the default tuning configuration for FXS and FXO voice ports, use the following commands beginning in privileged EXEC mode:

	Command	Task	Valid Entries	Default Values
Step 1	configure terminal	Enter the global configuration mode.
Step 2	voice-port slot-number/port	Identify the voice port you want to configure, and enter the voice port configuration mode.
Step 3	input gain value	Specify (in dB) the amount of gain to be inserted at the receiver side of the interface.	-6 to 14 dB	0 dB
Step 4	output attenuation value	Specify (in dB) the amount of attenuation at the transmit side of the interface.	0 to 14 dB	0 dB
Step 5	echo-cancel enable	Enable echo-cancellation of voice that is sent out of the interface and received back on the same interface.
Step 6	echo-cancel coverage value	Adjust the size (in milliseconds) of the echo-cancel.	8, 16, 24, and 32 ms	16 ms
Step 7	non-linear	Enable nonlinear processing, which shuts off any signal if no near-end speech is detected. (Nonlinear processing is used with echo-cancellation.)
Step 8	timeouts initial seconds	Specify the number of seconds the system will wait for the caller to input the first digit of the dialed digits.	0 to 120 sec	10 sec
Step 9	timeouts interdigit seconds	Specify the number of seconds the system will wait (after the caller has input the initial digit) for the caller to input a subsequent digit.	0 to 120 sec	10 sec
Step 10	timing digit milliseconds	If the voice-port dial type is DTMF, configure the DTMF digit signal duration.	50 to 100 ms	100 ms
Step 11	timing inter-digit milliseconds	If the voice-port dial type is DTMF, configure the DTMF inter-digit signal duration.	50 to 500 ms	100 ms
Step 12	timing pulse-digit milliseconds	(FXO ports only) If the voice-port dial type is pulse, configure the pulse digit signal duration.	10 to 20 ms	20 ms
Step 13	timing pulse-inter-digit milliseconds	(FXO ports only) If the voice-port dial type is pulse, configure the pulse inter-digit signal duration.	100 to 1000 ms	500 ms

Note After you change any voice-port command, we recommend that you cycle the port by using the shutdown and no shutdown commands.

To configure E&M voice ports, use the following commands beginning in privileged EXEC mode:

	Command	Required / Optional	Task
Step 1	configure terminal	Required	Enter the global configuration mode.
Step 2	voice-port slot-number/port	Required	Identify the voice port you want to configure, and enter the voice port configuration mode.
Step 3	dial-type {dtmf | pulse}	Required	Select the appropriate dial type for out-dialing.
Step 4	signal {wink-start | immediate | delay-dial}	Required	Select the appropriate signal type for this interface.
Step 5	cptone {australia | brazil | china | finland | france | germany | japan | northamerica | unitedkingdom}	Required	Select the appropriate voice call progress tone for this interface.
Step 6	operation {2-wire | 4-wire}	Required	Select the appropriate cabling scheme for this voice port.
Step 7	type {1 | 2 | 3 | 5}	Required	Select the appropriate E&M interface type. Type 1 is for the following lead configuration: E---output, relay to ground M---input, referenced to ground Type 2 is for the following lead configuration: E---output, relay to SG M---input, referenced to ground SB---feed for M, connected to -48V SG---return for E, galvanically isolated from ground Type 3 is for the following lead configuration: E---output, relay to ground M---input, referenced to ground SB---connected to -48V SG---connected to ground Type 5 is for the following lead configuration: E---output, relay to ground M---input, referenced to -48V.
Step 8	impedance {600c | 600r | 900c | complex1 | complex2}	Required	Specify a terminating impedance for an E&M voice port. The impedance value selected must match the specifications from the telephony system to which this voice port is connected.
Step 9	connection plar string	Optional	Specify the private line auto ringdown (PLAR) connection if this voice port is used for a PLAR connection. The string value specifies the destination telephone number.
Step 10	music-threshold number	Optional	Specify the threshold (in dB) for on-hold music. Valid entries are from -70 to -30 dB. The default is -38 dB.
Step 11	description string	Optional	Attach descriptive text about this voice-port connection.
Step 12	comfort-noise	Optional	Specify that background noise is generated.

• Pick up the handset of an attached telephony device, and listen for a dial tone.

• Check for DTMF detection if you have a dial tone. If the dial tone stops when you dial a digit, the voice port is configured properly.

• Use the show voice-port command to verify that the data configured is correct.

• Ping the associated IP address to confirm connectivity. If you cannot ping your destination, refer to the Network Protocols Configuration Guide, Part 1 for Cisco IOS Release 12.0T.

• Use the show voice-port command to make sure that the port is enabled. If the port is offline, use the no shutdown command.

• If you have configured E&M interfaces, make sure that the values pertaining to your specific PBX setup, such as timing and type, are correct.

• Make sure the VICs are correctly installed. For more information, refer to the <Emphasis>Cisco WAN Interface Cards Hardware Installation Guide.

If you need to change the default tuning configuration for E&M voice ports, use the following commands beginning in privileged EXEC mode:

	Command	Task	Valid Entries	Default Values
Step 1	configure terminal	Enter the global configuration mode.
Step 2	voice-port slot-number/port	Identify the voice port you want to configure, and enter the voice port configuration mode.
Step 3	input gain value	Specify (in dB) the amount of gain to be inserted at the receiver side of the interface.	-6 to 14 dB	0 dB
Step 4	output attenuation value	Specify (in dB) the amount of attenuation at the transmit side of the interface.	0 to 14 dB	0 dB
Step 5	echo-cancel enable	Enable echo-cancellation of voice that is sent out of the interface and received back on the same interface.
Step 6	echo-cancel coverage value	Adjust the size (in milliseconds) of the echo-cancel.	8, 16, 24, and 32 ms	16 ms
Step 7	non-linear	Enable nonlinear processing, which shuts off any signal if no near-end speech is detected. (Nonlinear processing is used with echo-cancellation.)
Step 8	timeouts initial seconds	Specify the number of seconds the system will wait for the caller to input the first digit of the dialed digits.	0 to 120 sec	10 sec
Step 9	timeouts interdigit seconds	Specify the number of seconds the system will wait (after the caller has input the initial digit) for the caller to input a subsequent digit.	0 to 120 sec	10 sec
Step 10	timing clear-wait milliseconds timing delay-duration milliseconds timing delay-start milliseconds timing dial-pulse min-delay milliseconds timing digit milliseconds timing inter-digit milliseconds timing pulse pulses-per-second timing pulse-inter-digit milliseconds timing wink-duration milliseconds timing wink-wait milliseconds	Specify timing parameters for each of these commands.	200 to 2000 ms 100 to 5000 ms 20 to 2000 ms 0 to 5000 ms 50 to 100 ms 50 to 500 ms 10 to 20 pps 100 to 1000 ms 100 to 400 ms 100 to 5000 ms

Note After you change any voice-port command, we recommend that you cycle the port by using the shutdown and no shutdown commands.

	Command	Task
Step 1	dial-peer voice number voip	Enter the dial peer configuration mode to configure a VoIP dial peer.
Step 2	acc-qos [best-effort | controlled-load | guaranteed-delay]	Generate an SNMP event if the QoS for a dial peer drops below a specified level.

Note RSVP reservations are only one-way. If you configure RSVP, the VoIP dial peers on either side of the connection must be configured for RSVP.

	Command	Task
Step 1	dial-peer voice number voip	Enter the dial peer configuration mode to configure a VoIP dial peer.
Step 2	codec [g711alaw | g711ulaw | g729r8 | g729r8 pre-ietf ]	Specify the desired voice coder rate of speech.

The default for the codec command is g729r8; normally, the default configuration for this command is the most desirable. However, if you are operating on a high bandwidth network and voice quality is of the highest importance, you should configure the codec command for g711alaw or ulaw. Using this value results in better voice quality, but it also requires higher bandwidth requirements for voice.

For example, to specify a CODEC rate of g711alaw for VoIP dial peer 108, enter the following:

Router(config)# dial-peer voice 108 voip
Router(config-dial-peer)# codec g711alaw

Note Prior to Cisco IOS Release 12.0(5)T, g729r8 is implemented in the pre-IETF format, thereafter it is implemented in the standard IETF format. Whenever new images (Release 12.0(5)T or later) interoperate with older versions of VoIP (when the g729r8 codec was not compliant with the IETF standard), users can hear garbled voices and ringback on either side of the connection. To avoid this problem, configure the dial peers with the g729r8 pre-ietf argument.

	Command	Task
Step 1	dial-peer voice number voip	Enter the dial peer configuration mode to configure a VoIP dial peer.
Step 2	vad	Disable the transmission of silence packets .

The default for the vad command is enabled; normally, the default configuration for this command is the most desirable. If you are operating on a high bandwidth network and voice quality is of the highest importance, you should disable VAD. Using this value results in better voice quality, but it also requires higher bandwidth requirements for voice.

For example, to enable VAD for VoIP dial peer 108, enter the following:

Router(config)# Dial-peer voice 108 voip
Router(config-dial-peer)# vad

In Cisco IOS Release 12.0T, Frame Relay traffic shaping is not compatible with RSVP. We suggest one of the following workarounds:

• Provision the Frame Relay PVC to have the CIR equal to the port speed.

• Use Generic Traffic Shaping with RSVP.

interface Serial0
 mtu 300
 no ip address
 encapsulation frame-relay
 no ip route-cache
 no ip mroute-cache
 fair-queue 64 256 1000
 frame-relay ip rtp header-compression
 
interface Serial1 point-to-point
 mtu 300
 ip address 40.0.0.7 255.0.0.0
 ip rsvp bandwidth 48 48
 no ip route-cache
 no ip mroute-cache
 bandwidth 64
 traffic-shape rate 32000 4000 4000
 frame-relay interface-dlci 16
 frame-relay ip rtp header-compression
 

In this configuration example, the main interface is configured as follows:

• MTU size is 300 bytes.

• No IP address is associated with this serial interface. The IP address must be assigned for the subinterface.

• Encapsulation method is Frame Relay.

• Fair-queuing is enabled.

• IP RTP header compression is enabled.

The subinterface is configured as follows:

Note When traffic bursts over the CIR, the output rate is held at the speed configured for the CIR (for example, traffic will not go beyond 32 kbps if CIR is set to 32 kbps).

For more information about configuring Frame Relay for VoIP, refer to the "Configuring Frame Relay" chapter in the Wide-Area Networking Configuration Guide for Cisco IOS Release 12.0T.

Step 2 Click the Audio tab.

Step 3 Select the "Calling a telephone using NetMeeting" check box.

Step 4 Enter the IP address of your router in the IP address field.

Step 5 Under General, click Advanced.

Step 6 Select the "Manually configured compression settings" check box.

Step 7 Select the CODEC value CCITT ulaw 8000Hz.

Step 8 Click the Up button until this CODEC value is at the top of the list.

Step 9 Click OK to exit.