The DSP Voice Quality Statistics in DLCX Messages feature provides a way to trace a Media Gateway Control Protocol (MGCP) call between a Cisco PGW 2200 Softswitch and the Cisco IOS gateway by including the MGCP call ID and the DS0 and digital signal processor (DSP) channel ID in call-active and call-history records.
The voice quality statistics are sent as part of the MGCP Delete Connection (DLCX) message. By correlating an MGCP call on the Cisco PGW 2200 Softswitch with a call record on the gateway, you can understand and debug additional statistics from the DSP for problems related to voice quality.
Your software release may not support all the features documented in this module. For the latest feature information and caveats, see the release notes for your platform and software release. To find information about the features documented in this module, and to see a list of the releases in which each feature is supported, see the Feature Information Table at the end of this document. Use Cisco Feature Navigator to find information about platform support and Cisco software image support. To access Cisco Feature Navigator, go to
http://www.cisco.com/go/cfn. An account on Cisco.com is not required.
Prerequisites for DSP Voice Quality Statistics in DLCX Messages
You must be using Cisco PGW 2200 version 9.4.1 or a later version with a patch level higher than CSCOgs008/CSCOnn008.
Restrictions for DSP Voice Quality Statistics in DLCX Messages
When the Secure Real-Time Transfer Protocol (SRTP) is enabled, the DLCX message will not report voice quality statistics. The following lines will be omitted for SRTP calls:
DSP/Endpoint Configuration (EC)
DSP/MOS K-Factor Statistics (KF)
DSP/Concealment Statistics (CS)
DSP/R-Factor Statistics (RF)
DSP/User Concealment (UC)
Information About DSP Voice Quality Statistics in DLCX Messages
Cisco PGW 2200
call agent (or
media gateway controller) and
are industry standard terms used to describe the network element that provides call control functionality to telephony and packet networks. The Cisco PGW 2200 Softswitch functions as a call agent or softswitch in “call control mode.”
All voice quality parameters for Cisco IOS Release 12.4(4)T and later releases are supported only on the Cisco PGW 2200 call agent.
A public switched telephone network (PSTN) gateway provides an interface between traditional Signaling System No. 7 (SS7) or non-SS7 networks and networks based on MGCP, H.323, and Session Initiation Protocol (SIP), which include signaling, call control, and time-division multiplexing/IP (TDM/IP) gateway functions. The Cisco PGW 2200 Softswitch, coupled with Cisco media gateways, functions as a PSTN gateway.
There is a significant performance degradation on the Cisco PGW 2200 if all connected gateways have the DSP Voice Quality Statistics in DLCX Messages feature enabled. Enabling voice quality statistics on the gateway should only be performed by Cisco personnel.
The Cisco PGW 2200 Softswitch, in either signaling or call control mode, provides a robust, carrier-class interface between PSTN and IP-based networks. Interworking with Cisco media gateways, the Cisco PGW 2200 Softswitch supports a multitude of applications and networks, including the following:
Application service provider (ASP) termination
Centralized routing and billing for clearinghouse of IP-based networks
International and national transit networks
Managed business voice applications
Managed voice VPNs
Network clearinghouse applications
PSTN access for hosted and managed IP telephony
PSTN access for voice over broadband networks
Residential voice applications
MGCP defines the call control relationship between call agents (CAs) and VoIP gateways that translate audio signals to and from the packet network. CAs are responsible for processing the calls.
An MGCP gateway handles the translation between audio signals and the packet network. The gateways interact with a CA, also called a media gateway controller (MGC), which performs signal and call processing on gateway calls. MGCP uses endpoints and connections to construct a call.
Endpoints are sources of or destinations for data and can be physical or logical locations in a device. Connections can be either point-to-point or multipoint. The gateway can be a Cisco router, an access server, or a cable modem, and the CA is a server from a third-party vendor.
The Cisco PGW 2200
Softswitch can capture voice quality statistics sent from MGCP-controlled media
gateways and can propagate the statistics into call detail records (CDRs) at
the end of each call. The Cisco AS5x00 media gateways send voice quality
statistics to the Cisco PGW 2200 Softswitch.
Most voice quality
statistics are available from the DSP and are controlled using RTP Control
Protocol (RTCP) report interval statistics polling. The mean and maximum values
are calculated by Cisco IOS software-based polling, which results in additional
CPU load for each call. The additional CPU load can be controlled by
configuring polling interval by using the
The playout delay,
playout error, and DSP receive and transmit statistics are automatically polled
periodically. Polling for the voice quality statistics, level, and error
parameters can be added. For logging the voice quality statistics using syslog,
the existing VoIP gateway accounting has been extended. For more information
about statistics polling, see the
ip rtcp report
interval command in the
Table 1 Voice Quality Statistics for
Cisco IOS Release 12.4(4)T and Later Releases
Cisco AS5350XM, Cisco AS5400, Cisco AS5400HPX, Cisco 5400XM, and Cisco AS5850
with an NPE60 or NPE108 universal port feature card.
2800 series and Cisco 3800 series integrated services routers with PVDM2
VG224 voice gateway
IAD2430 series integrated access devices.
2600XM, Cisco 2691, Cisco 3700 series access routers, and Cisco 2811, Cisco
2821, Cisco 2851, Cisco 3800 series integrated services routers with the
following network modules: NM-HDV2, NM-HDV2-1T1/E1, NM-HD-1V, NM-HD-2V,
2821, Cisco 2851, Cisco 3825, and Cisco 3845 with the EVM-HD-8FXS/DID module.
The DSP Voice Quality Statistics in DLCX Messages feature is part of the Cisco quality of service (QoS) technology. QoS is the ability of a network to provide better service to selected network traffic over various technologies, including ATM, Ethernet and 802.1 networks, Frame Relay, SONET, and IP-routed networks that may use any or all of these underlying technologies.
QoS provides the following benefits:
Control over bandwidth, equipment, and wide-area facilities—As an example, you can limit the bandwidth consumed over a backbone link by FTP or limit the queueing of an important database access.
More efficient use of network resources—Network analysis management and accounting tools enable you to know what your network is being used for and ensure that you are servicing the most important traffic to your business.
Ability to customize services—QoS enables ISPs to offer carefully customized grades of service differentiation to their customers.
Coexistence of mission-critical applications—Cisco QoS technologies ensure that bandwidth and minimum delays required by time-sensitive multimedia and voice applications are available and that other applications using the link get their fair service without interfering with mission-critical traffic.
Foundation for a fully integrated network—Cisco QoS technologies fully integrate a multimedia network, for example, by implementing weighted fair queueing (WFQ) to increase service predictability and by implementing IP precedence signaling to differentiate traffic. In addition, the availability of Resource Reservation Protocol (RSVP) allows you to take advantage of dynamically signaled QoS.
To deliver QoS across a network that comprises heterogeneous technologies (for example, IP, ATM, LAN switches), basic QoS architecture consists of the following three components:
QoS within a single network element (for example, queueing, scheduling, and traffic shaping tools).
QoS signaling techniques for coordinating end-to-end QoS between network elements.
QoS policy, management, and accounting functions to control and administer end-to-end traffic across a network.
The following parameters describe the configuration of a VoIP endpoint. You can define these parameters and they are useful for debugging and logging purposes because they capture the state of the endpoint.
CI—Codec ID. A string or a number that identifies the voice codec which is currently used in the call.
FM—Frame size. Native frame size, in milliseconds (ms), of the selected codec. An example of a frame size and codec combination is G.729a/30ms. For the G.711 codec, the frame size is a value that you can define in the voice dial peer. For example, G.711 at 80 bytes gives 10 ms per frame. G.711 at 240 bytes gives 30 ms per frame.
FP—Frames per packet. Number of codec speech frames encapsulated into a single Real-time Transport Protocol (RTP) packet. Typical values are 1, 2, and 3. Packing lower number of frames per packet results in lower efficiency of IP bandwidth usage. The tradeoff is lower delays and higher robustness of the network.
VS—Voice Activity Detection (VAD)-enabled flag. VAD is enabled when VS has a value of one. It results in compression of silent periods leading to reduced or zero packets per second. VAD is disabled when VS has a value of zero. It results in the transmission of continuous packets per second irrespective of active or silent periods on the transmission path.
GT—Transmission gain factor (linear). Digital gain multiplier applied to the transmission on the signal path from the PSTN toward the network. GT is applied at the echo canceller
Sout port. A gain factor of less than one indicates a loss pad.
GR—Reception gain factor (linear). Digital gain multiplier applied to reception on the signal path from the network toward the PSTN. GR is applied at the echo canceller
Rin port. A gain factor of less than one indicates a loss pad.
JD—Jitter buffer mode. It consists of the following modes:
Adaptive mode = 1
Fixed mode (no timestamps) = 2
Fixed mode (with timestamps) = 3
Fixed mode (with passthrough) = 4
JN—Jitter buffer nominal playout delay. Size of the jitter buffer in milliseconds. An adaptive jitter buffer tries to make the playout delay equal to the nominal (desired) delay when the observed jitter is small enough to allow this adjustment. For a fixed-mode jitter buffer, the nominal setting is the constant playout delay itself.
JM—Minimum playout delay. Minimum playout delay setting for an adaptive-mode jitter buffer. The playout delay never goes below the minimum playout setting even if the observed jitter is zero. This setting is not used for a fixed-mode jitter buffer because the playout delay is fixed and constant at the nominal setting.
JX—Maximum playout delay. Sets the limit for increasing the playout delay of an adaptive-mode jitter buffer. An adaptive buffer increases when the jitter is higher than the instantaneous playout delay value.
K-factor is an endpoint mean opinion score (MOS) estimation algorithm defined in the ITU standard P.VTQ. It is a general estimator of the mean value of a perceptual evaluation of speech quality (PESQ) population for a specific impairment pattern.
The ITU standard P.862 defines and describes the PESQ as an objective method for end-to-end speech quality assessment of narrow band telephone networks and speech codecs.
Mean opinion score (MOS) is associated with the output of a well-designed listening experiment. All MOS experiments use a five-point PESQ scale as defined in the ITU standard P.862.1. The MOS estimate is inversely proportional to frame loss density. Clarity decreases as more frames are lost or discarded at the receiving end.
K-factor represents a weighted estimate of average user annoyance due to distortions caused by effective packet loss such as dropouts and warbles. It does not estimate the impact of delay-related impairments such as echo. It is an estimate of listening quality (MOS-LQO) rather than conversational quality (MOS-CQO), and measurements of average user annoyance range from 1 (poor voice quality) to 5 (very good voice quality).
K-factor is trained or conditioned by speech samples from numerous speech databases, where each training sentence or network condition associated with a P.862.1 value has a duration of eight seconds. For more accurate scores, K-factor estimates are generated for every 8 seconds of active speech.
K-factor and other MOS estimators are considered to be secondary or derived statistics because they warn a network operator of frame loss only after the problem becomes significant. Packet counts, concealment ratios, and concealment second counters are primary statistics because they alert the network operator before network impairment has an audible impact or is visible through MOS.
KF—K-factor MOS-LQO estimate (instantaneous). Estimate of the MOS score of the last 8 seconds of speech on the reception signal path. If VAD is active, the MOS calculation is suspended during periods of received silence to avoid inflation of MOS scores for calls with higher silence fractions.
AV—Average K-factor score. Running average of scores observed since the beginning of a call.
MI—Minimum K-factor score. Minimum score observed since the beginning of a call, and represents the worst sounding 8-second interval.
BS—Baseline (maximum) K-factor score. K-factor score that can be obtained for the defined codec.
NB—Number of bursts. Number of burst loss events after a call is started. A burst loss is a contiguous run of concealment events of length greater than one.
FL—Average frame loss count. Total number of frame losses and concealment events observed after starting a call. The ratio of FL/NB provides the mean burst length in frames. The total concealment duration of the call is provided by the parameter
NW—Number of windows. Total number of K-factor windows observed after starting a call. The number of windows is directly proportional to the duration of a call.
VR—Version ID. Version number that identifies a specific K-factor MOS score.
DSP/CS measures packet (frame) loss and its effect on voice quality in an impaired network. The parameters for concealment statistics are as follows:
CR—Concealment ratio (instantaneous). An interval-based average concealment rate, and is the ratio of concealment time over speech time for the last 3 seconds of active speech. When VAD is enabled, calculation of the concealment ratio is suspended during periods of speech silence. During this suspension, it may take more than 3 seconds for a new value to be generated.
AV—Average CR. Average of all CR reports after starting a call.
MX—Maximum CR. The maximum concealment ratio observed after starting a call.
CS—Concealed time. The duration of time in seconds during which some concealment is observed.
CT—Total concealment time. The total duration of time in milliseconds during which concealment is observed after starting a call.
TT—Total speech time. The duration of time in milliseconds during which active speech is observed after starting a call.
OK—Ok time. The duration of time in seconds during which no concealment is observed.
SC—Severely concealed time. The duration of time in seconds during which a significant amount of concealment is observed. If the concealment observed is usually greater than 50 milliseconds or approximately five percent, it is possible that the speech is not very audible.
TS—Concealment threshold. The threshold in milliseconds used to determine a second as severely concealed. The threshold for concealed seconds is 0 milliseconds, and for severely concealed seconds is 50 milliseconds.
The R-factor helps in planning voice transmission. In ITU standards G.107 and G.113, the R-factor is defined as follows:
R = Ro - Is - Id - Ie-eff + A
The parameters for the R-Factor are as follows:
Ro is based on the signal-to-noise ratio.
Is is the simultaneous impairment factor and includes the overall loudness rating.
Id is the delay impairment factor and includes talker (Idte) and listener (Idle) echos, and delays (Idd).
Ie-eff is the equipment impairment factor and includes packet losses and the types of codecs.
A is the advantage factor.
ML—R-factor MOS listening quality objective. It reflects only packet loss and codec-related impairments and does not include delay effects.
R1—R-factor LQ profile 1.
R2—R-factor LQ profile 2.
IF—Effective codec impairment (Ie_eff).
IE—Codec baseline score (Ie). The tabulated baseline codec impairment factor.
BL—Codec baseline (Bpl). The packet loss robustness factor for the codec being used.
R0—R0 (default). The nominal value at which the signal-to-noise ratio is considered nominal.
The parameters for user concealment are as follows:
U1—User concealment seconds 1 count (UCS1)
U2—User concealment seconds 2 count (UCS2)
T1—UCS1 threshold in milliseconds
T2—UCS2 threshold in milliseconds
The parameters for delay statistics are as follows:
RT—Round trip delay
ED—End system delay
How to Configure DSP Voice Quality Statistics in DLCX Messages
Configuring DSP Voice Quality Statistics in DLCX Messages
To configure voice quality statistics reporting for MGCP, use the following commands beginning in user EXEC mode.
Priority2 is similar to using the
all keyword when the output contains the following parameters: DSP/TX, DSP/RX, DSP/PD, DSP/PE, DSP/LE, DSP/ER, DSP/IC, DSP/EC, DSP/KF, DSP/CS, DSP/RF, DSP/UC, and DSP/DL.
Exits global configuration mode and enters privileged EXEC mode.
What to Do Next
Use the following Troubleshooting Tips if you did not get the expected results after configuring voice quality statistics reporting for MGCP. See the
"Troubleshooting Tips" section for additional guidelines.
Cisco Support and Documentation website provides online resources to download
documentation, software, and tools. Use these resources to install and
configure the software and to troubleshoot and resolve technical issues with
Cisco products and technologies. Access to most tools on the Cisco Support and
Documentation website requires a Cisco.com user ID and password.
Feature Information for DLCP Voice Quality Statistics in DLCX Messages
The following table provides release information about the feature or features described in this module. This table lists only the software release that introduced support for a given feature in a given software release train. Unless noted otherwise, subsequent releases of that software release train also support that feature.
Use Cisco Feature Navigator to find information about platform support and Cisco software image support. To access Cisco Feature Navigator, go to
www.cisco.com/go/cfn. An account on Cisco.com is not required.
Table 2 Feature Information for DLCP Voice Quality in DLCX Messages
DSP Voice Quality Statistics in DLCX Messages
The DSP Voice Quality Statistics in DLCX Messages feature provides a way to trace a Media Gateway Control Protocol (MGCP) call between a Cisco PGW 2200 and the Cisco IOS gateway by including the MGCP call ID and the DS0 and digital signal processor (DSP) channel ID in call-active and call-history records.
In Cisco IOS Release 12.4(4)T, new voice quality parameters were introduced.
The following commands were introduced or modified:
mgcp voice-quality stats.