HD Voice Communication - Coding and Enhancement

We recommend you to submit your preliminary or firm registration at least 4 weeks before course start to ensure a seat on the course.

TECHNOLOGY FOCUS 
The focus of this course is on speech-audio coding and advanced signal processing algorithms for GSM, UMTS or LTE mobile phones, digital hearing aids, and human-machine interfaces. Within the evolution of these systems, the improvement of the speech-audio quality will remain one of the most important objectives to mitigate physical constraints and technological limitations. A comprehensive understanding of fundamental algorithms, standards, applications, and trends is provided. Conditions and solutions are presented with regards to audio-bandwidth limitation, bit-rate restrictions, interference by acoustic background noise, reverberation, acoustic echo signals, and residual transmission errors. 

COURSE CONTENT 
The course covers theory, practice, standards and trends of speech-audio communication, including state-of-the art solutions, the on-going conversion of networks and terminals to HD voice as well as emerging concepts for spatial HD audio communication:

Speech-Audio Coding standards: ETSI, 3GPP, ITU, and MPEG

• Coding concepts for spatial HD audio communication
• Error Concealment & Soft Decision Source Decoding
• Audio Bandwidth Extension: With and without side information
• Noise Reduction & Dereverberation: Single- and Multi-Microphone Techniques
• Acoustic Echo Cancellation: Time- and frequency-domain, adaptive postfiltering

The coding and processing techniques are demonstrated by audio examples.

Figure: Digital Speech Transmission and Enhancement

Monday
Speech-Audio CODING 
The first part of the course deals with speech-audio coding. State-of-the-art concepts are discussed. The signal processing aspects of quantization, differential waveform coding, linear prediction, Code Exited Linear Prediction (CELP) and Transform Coding are explained.

Speech Production Model

• Speech Production
• Digital Filter Structures for Speech Production
• Psycho-Acoustics

Linear Prediction

• Vocal Tract Models and Short-Term Prediction
• Optimum Prediction
• Spectral Flatness Measure
• Block-Adaptive Linear Prediction
• Levinson/Durbin Algorithm
• Long-Term Prediction

Quantization

• Uniform and Non-uniform Quantization
• Optimal Quantization
• Adaptive Quantization
• Vector Quantization

Speech-Audio Coding

• Model-Based Predictive Coding
• Adaptive Differential Pulse Code Modulation (ADPCM)
• Noise Shaping Open Loop and Closed Loop Prediction
• Code Excited Linear Prediction (CELP)
• Quantization and Line Spectral Frequencies (LSF)
• Transform- and Subband Coding
• Adaptive Bit Allocation
• Audio Examples: Quantization, Coding, Noise Shaping

Tuesday
Post Filtering

• Short-term Post Filter
• Long-term Post Filter
• Tilt Compensation

Quality Assessment

• Mean Opinion Score (MOS)
• Modulated Noise Reference Unit (MNRU, CCITT)
• Objective Quality Measures (PESQ)

Coding Standards 
The most relevant speech-audio codec standards for real-time communication are discussed and demonstrated by audio examples:

• ITU: G.721/G.726, G.722, G.721, G.723, G.728, G.729, G.729.1, G.711.1, G.718
• ETSI/3GPP: Full Rate, Half Rate, Enhanced Full Rate, AMR, AMR-WB, AMR-WB+
• MPEG: AAC-LD, AAC-ELD
• Proprietary: SILK, CELT

ERROR CONCEALMENT and SOFT DECISION SOURCE DECODING
Wireless speech transmission systems usually include channel coding for error protection. 
However, due to temporarily adverse channel conditions quite frequently residual bit and frame errors remain. The negative effects of these errors can be reduced by error concealment, exploiting both residual source redundancy and information about the instantaneous quality of the transmission channel:

• Frame Substitution and Standard Solutions (GSM)
• Soft-bits and Log. Likelihood Values
• Soft-Decision Speech Decoding by Parameter Estimation
• A Priori Knowledge and A Posteriori Probabilities
• Graceful Degradation by Soft Decoding
• Joint and Iterative Source-Channel (De-) Coding
• Audio Examples: PCM, ADPCM, GSM

Wednesday
BANDWIDTH EXTENSION 
In the long run, the audio bandwidth of the telephone networks and terminals will be extended to 7 kHz wideband transmission. This will require new codecs at both sides of the transmission link. In the probably very long transition period many terminals have not yet been equipped with the wideband capability. In this situation, the quality of the received narrowband speech (3.4 kHz) may be improved by means of artificial bandwidth extension:

• Source Filter Model
• Extension of the Excitation Signal
• Extension of the Spectral Envelope
• Statistical Estimation Based on a Markov State Model
• Bandwidth Extension with Side Information
• Implementation and Performance Evaluation

NOISE REDUCTION, DEREVERBERATION AND BEAMFORMING
If the signal is degraded by acoustic background noise, dereverberation and a loudspeaker signal, various speech enhancement methods can be applied prior to the speech encoding and transmission. We will discuss state-of-the-art algorithms for noise reduction, dereverberation and evaluate new proposals such as super-Gaussian speech models and psycho-acoustic aspects. Noise suppression schemes using only one single microphone or several microphones are presented.

Single and Dual Channel Noise Reduction

• Wiener Filter
• Speech Enhancement in the DFT Domain
• Noise Estimation Techniques: Minimum Statistics, MMSE
• More Sophisticated Suppression Rules
• Conditional MMSE- and MAP-Estimation
• Estimation of Complex DFT-Coefficients
• Estimation of Real Valued DFT-Amplitudes
• Estimation with Super-Gaussian Models
• Noise Suppression Exploiting Masking
• Soft Weighting
• Dual Channel Noise Cancellation
• Coherence Function and Theoretical Limits
• Dual Channel Noise Suppression
• Noise Suppression using Small Microphone Arrays
• Audio Examples

Thursday
Multi-channel Noise Reduction

• Spatial Sampling of Sound Fields
• Beamforming
• Performance Measures
• Fixed Beamformers
• Multi-channel Wiener Filter and Postfilter
• Adaptive Beamformers
• Generalized Side-lobe Canceller

ACOUSTIC ECHO CONTROL 
The key algorithms for acoustic echo control used for hands-free communication 
are explained, especially for the combination of echo cancellation 
with adaptive post-filtering.

• LMS and NLMS Time Domain Cancellation
• Convergence Analysis and Control
• Echo Cancellation and Postfiltering
• Joint Residual Echo Cancellation and Noise Reduction
• Frequency Domain Method and Block Processing
• Additional Measure
• Stereophonic Acoustic Echo Control
• Audio Examples

Said about the course from previous participants:
"Good overview of all topics."
"The course gave deep theoretical background."
"Complete overview of Speed Transmission, Coding and Noise Reduction /EC."
"Competent teacher and a very industry oriented course."