Speech Quality of VoIP

Speech Quality of VoIP

Speech Quality of VoIP addresses speech quality of VoIP networks, whilst addressing the implications for the design of VoIP networks and systems. The book bridges the gap between the technical network-world and the (psychoacoustic) world of quality perception. Starting from the network designer's point of view, the different characteristics of the network are addressed, and then linked to features perceived by users. Since quality can be viewed as the result of a comparison of perceived features with desired features, not only the perception of features determines quality, but also the features a user desires of a system. The book addresses both of these aspects.

Speech Quality of VoIP is an essential guide to assessing the speech quality of VoIP networks, whilst addressing the implications for the design of VoIP networks and systems. This book bridges the gap between the technical network-world and the psychoacoustic world of quality perception. Alexander Raake's unique perspective combines awareness of the technical characteristics of VoIP networks and original research concerning the perception of speech transmitted across them.

Starting from the network designer's point of view, the different characteristics of the network are addressed, and then linked to features perceived by users. This book provides an overview of the available knowledge on the principal, relevant aspects of speech and speech quality perception, of speech quality assessment, and of transmission properties of telephone and VoIP networks, and of the related perceptual features and resulting speech quality. Discussing new research into the specific time-varying degradations VoIP brings along, but also the considerable potential of quality improvement to be achieved with wideband speech transmission, Alexander Raake demonstrates how network and service characteristics impact on the users perception of quality.

Speech Quality of VoIP:

- Offers an insight into speech quality of VoIP from a user's perspective.

- Presents an overview of different modelling approaches and a parametric network-planning model for quality prediction in VoIP networks.

- Draws on innovation new research on the quality degradation characteristic of VoIP.

- Explains in detail how telephone speech quality can be enhanced with VoIP's wideband speech transmission capability.

- Assesses the vast collection of references into the technical and scientific literature related to VoIP quality.

- Illustrates concepts throughout with mathematical models, algorithms and simulations.

Speech quality of VoIP is the definitive guide for researchers, engineers and network planners working in the field of VoIP, Quality of Service, and speech communication processing in telecommunications. Advanced undergraduate and graduate students on telecommunication and networking courses will also find this text an invaluable resource.

---

Inhaltsangabe

Preface

Introduction

Overview of the Book

1. Speech Quality in Telephony

1.1 Speech

1.1.1 Speech Acoustics

1.1.2 Speech Perception

1.2 Speech Quality

1.2.1 Definition of Quality

1.2.2 Speech Quality Assessment

1.2.3 Quality Elements

1.2.4 Speech Quality and Quality of Service

2. Speech Quality Measurement Methods

2.1 Auditory Methods

2.1.1 Utilitarian Methods

2.1.2 Analytical Methods

2.2 Instrumental Methods

2.2.1 Signal-based Models

2.2.2 Parameter-based Models

2.2.3 Monitoring Models

2.3 Speech Quality Measurement Methods: Summary

3. Quality Elements and Quality Features of VoIP

3.1 Speech Transmission Using the Internet Protocol

3.2 Overview of Quality Elements

3.3 Quality Elements and Related Features

3.3.1 Voice Activity Detection (VAD)

3.3.2 Codecs

3.3.3 Jitter

3.3.4 Delay

3.3.5 Packet Loss

3.3.6 Talker Echo

3.3.7 Listener Echo

3.3.8 Echo Cancellers

3.3.9 Loudness

3.3.10 Noise

3.3.11 Noise Suppression

3.3.12 Linear Distortion and Wide-band Transmission

3.3.13 User Interfaces

3.4 Quality Dimensions

3.5 Combined Elements and Combined Features

3.6 Listening and Conversational Features

3.7 Desired Nature

3.8 Open Questions

3.8.1 From Quality Elements to Quality Features: The E-model and VoIP

3.9 Quality Elements and Quality Features of VoIP: Summary

4. Time-Varying Distortion: Quality Features and Modelling

4.1 Microscopic Loss Behavior

4.1.1 Random Packet Loss

4.1.2 Dependent Packet Loss

4.2 Macroscopic Loss Behavior

4.2.1 Macroscopic Loss with Segments of Random Loss

4.2.2 Macroscopic Loss with Segments of Dependent Loss

4.3 Interactivity

4.4 Packet Loss and Combined Impairments

4.4.1 Additivity and Multidimensional Feature Space

4.4.2 Microscopic Loss Behavior: Random Packet Loss

4.4.3 Macroscopic Loss Behavior: 3-state Markov Loss

4.4.4 Model: Packet Loss and Combined Impairments

4.5 Time-Varying Distortion: Summary

5. Linear Distortion & Wideband Speech

5.1 Wideband Speech: Improvement over Narrowband

5.2 Bandpass-Filtered Speech

5.2.1 Listening Test

5.2.2 Parametric Impairment Model

5.3 Desired Nature

5.3.1 User Interfaces

5.3.2 Content

5.3.3 Summary

6. Summary and Conclusions

A. Aspects of Parametric Packet Loss Descriptions

A.1 4-state Markov Chain: Sojourn in Bad and Good State

A.2 Impact of other Quality Elements on Packet Loss

A.2.1 Forward Error Correction (FEC)

A.2.2 Jitter and Jitter Buffer

B. Simulation of Quality Elements

B.1 PSTN and ISDN

B.1.1 Instrumental Verification

B.2 Mobile Networks

B.3 VoIP

B.3.1 Analog to ISDN Conversion

B.3.2 VoIP Gateways

B.3.3 IP Network Simulation

B.3.4 Instrumental Verification

B.4 Wideband Transmission

B.5 User Interfaces

B.5.1 Headsets

B.5.2 Wideband Handset and Hifi Phone

B.5.3 Bandpass Filters

B.5.4 Hands-Free Terminals (HFTs)

B.6 Test Rooms

B.7 Simulation of Quality Elements: Summary

C. Frequency Responses

C.1 Transmission Bandpass

C.1.1 Narrowband

C.1.2 Wideband

C.2 User Interfaces

C.2.1 Headsets

C.2.2 Wideband Handsets

D. Test Data Normalization and Transformation

D.1 Equivalent-QMethod

D.2 Method According to ITU-T Rec. P.833

D.3 Linear Transformation

E. E-model Algorithm

F. Interactive Short Conversation Test Scenarios (iSCTs)

F.1 Example

G. Auditory Test Settings and Results

G.1 GSM: SCT and LOT

G.2 2-State Markov Loss: LOT

G.3 Random Loss: CT

G.3.1 Test Set-up

G.3.2 Test Procedure

G.3.3 Subjects

G.3.4 Details on Selected Results

G.4 3-State Markov Loss: CT

G.4.1 Test Set-Up

G.4.2 Test Procedure and Test Subjects

G.4.3 Details on Selected Results

G.5 Speech Sound Quality and Content

H. Modelling-Details

H.1 Time-Varying Distortions

H.1.1 Macroscopic Loss Behavior

Bibliography