Copyright Telchemy Inc 2001 Embedded Passive Monitoring of Voice over IP Service Quality Alan Clark Telchemy, Inc Atlanta, GA [email protected]

Copyright Telchemy Inc 2001

Embedded Passive Monitoring of Voice over IP Service Quality

Alan Clark

Telchemy, IncAtlanta, GA

[email protected]


Embedded Passive Monitoring

IPNetwork

GatewayGateway

Monitoring Agent embedded into VoIP Gateway

QoSmetrics


Active vs Passive Testing

• Active Measurement– Establish looped back test call– Send recorded speech file– Compare sent and received files (PSQM/PESQ)

• Passive Monitoring– Measure characteristics of live calls– Determine Transmission Quality metric– Estimate Speech Quality metrics


The E Model

• “Mouth to ear” transmission quality measurement

• Produces an “R” factor typically in the range 50 (bad) -95 (good)

• R factor can be related to MOS score, Terminate Early (TME) etc.

• ITU G.107/ G.108 and ETSI ETR250


E Model

R = Ro - Is - Id - Ie + A

Base R value- Noise level

Impairments thatoccur simultaneouslywith speech- received speech level- sidetone level- quantization noise

Impairments thatare delayed withrespect to speech- talker echo- listener echo- round trip delay

Equipment ImpairmentFactor- CODEC- multiplexing effects

Advantage factor


E Model

Delay, measuredusing RTCP

R Factor

E Model

Ie

PacketLoss

Jitter

Codectype

LossModel

JitterModel

CodecModel


R Factor vs MOS

50

60

70

80

90

100

0.1 1 10

Percentage of users that terminate calls early

4.5

4.0

3.0

R Factor MOS


Impairments can vary with time

• Interaction of streams in IP network causes many impairments to vary with time

• Adaptive jitter buffer

• Call re-routing

• CODEC type changed during call


Measuring Packet Loss

ArrivingRTPpackets

Jitterbuffer

Late packetsdiscarded

Monitor jitter andpacket loss afterjitter buffer

CODEC

RTCP statisticsreport every 5 seconds


Burst vs average loss

Burst of packet lossZero packet loss

Non-bursty packet loss

Is not the same as


Effects of packet loss burst

TransmissionQuality

High

Low

4-5seconds

15-30seconds

User perceivedquality profile


Loss Model - Markov model

3Lost

1Rcvd

P13

P22

P11

P312

Rcvd

P23P32

4Lost

P14

P41

Burst state

Gap state


Frame loss event driven

Burst of packet lossZero packet loss

Lost( ) Lost( ) Lost( )


Loss Model - Markov model

3Lost

1Rcvd

P13

P22

P11

P312

Rcvd

P23P32

4Lost

P14

P41

Burst state

Gap state

Model parametersreconstructed at endof call


Loss Model - mapping loss to Ie

0

10

20

30

40

50

0 5 10 15

Packet Loss Rate

Ie (p

acke

t los

s)Curve is CODECdependant


Determining QoS metrics

1. Determine “good”and “bad” state Ie Factor



t = 5

t = 15


2. EstimateInstantaneousR Factor foreach state



t = 5

t = 15


2. EstimateInstantaneousR Factor foreach state

3. Determine average Ie


Measuring Delay

CODECCODEC

Accumulateframe

Encode

Transmission Jitterbuffer

Decode

RTCP exchange


Delay Model

0

10

20

30

0 100 200 300 400

175 mS “knee”

End to end delay (mS)

R FactorReduction


“Recency” effect

“Good” 4.3MOS“Bad” 1.8 MOS(3dB SNR)

MOS 3.82

MOS 3.28

MOS 3.18

Source AT&TT1A1.7/98-031

60 second call


Estimation of recency effect

Delay since last significant burst

Effects decay over 30-60 secondsAverage for call


Integration with VoIP SMS

VQmonSNMPSet, Get, Trap

End of callmsg (DRQ)

CDRServiceMgtSystem

NetworkMgtSystem


H.323/H.225 based reporting

H.225 Disengage Request (DRQ)

nonStandardData fieldRend INTEGER (0..100)Rav INTEGER (0..100)Rmin INTEGER (0..100)Rmax INTEGER (0..100)


Embedded Passive Monitoring

RTP stream CODEC

IPNetwork

CODEC

1 Measure Packet Loss Distribution, Jitter, Delay, CODEC type

2 Produce a transmission quality rating (R Factor) using extension to the E-Model

3 Map R Factor to an estimated MOS Score

Documents

Copyright Telchemy Inc 2001 Embedded Passive Monitoring of Voice over IP Service Quality Alan Clark Telchemy, Inc Atlanta, GA [email protected]