(Continuation of Question 7/15 and part of Question 5/15)

Background and justification

This Question deals with network-based speech enhancement devices, including electrical network echo control, acoustic network echo control, automatic level control devices, and speech enhancement devices for wireless networks.

The application of echo cancellers continues to evolve in existing PSTN, cellular, and emerging ATM and IP networks. These applications can bring new issues to echo canceller performance, and will require study to ensure that administrations have adequate assurance of transmission quality with echo cancellers complying with G.168. These issues may result in adding and improving the tests within G.168, and studying the issues arising from the deployment of echo cancellers to ATM and IP networks and new services within the PSTN/GSTN.

There is also a need to address equipment-related issues associated with echo cancellers such as; defining appropriate interfaces of echo cancellers, alarming requirements, power requirements, and the possible integration of automatic level control and other speech enhancement functions into echo cancellers.

The deployment of Automatic Level Control Devices is just beginning, and new issues are expected to arise as these devices are applied to existing and emerging networks. As a result, G.169 may need to be revised to ensure adequate transmission quality. Wireless echo cancellers and speech enhancement devices are being deployed in cellular networks to improve the performance of mobile applications. There is consequently a need for a new Recommendation to cover the performance requirements of these devices.

Responsibility under this Question includes the G.16x series of Recommendations and those aspects of the G.776 series that deal with the operation of echo cancellers.

Question

• What functional characteristics and requirements are necessary for network-based speech enhancement devices to provide good performance in existing GSTN and emerging ATM/IP and cellular networks?
• What requirements are necessary for the equipment-related aspects of network-based speech enhancement devices?
• What management interfaces should be recommended for speech enhancement equipment?

Study items

Study items to be considered include:

• The enhancement of objective tests to better correlate with subjective test results
• The effect of impairments, including level and PCM offset, poor natural echo return loss, noise, and possible phase roll on the performance of echo cancellers and speech enhancement devices
• Tests and requirements necessary to ensure good double-talk performance of echo cancellers. These tests may cover measurement of clipping level variations and echo during doubletalk
• The effect of tandem echo cancellers
• The effect of low bit rate coders and other non-linearities in the echo path of echo cancellers
• The performance of speech enhancement devices with facsimile and voice-band data signals, signalling tones, call-processing tones, speech, etc.
• The performance of speech enhancement devices in cellular and other wireless environments.
• The performance of the non-linear processor in an echo canceller
• Background noise transmission through speech enhancement devices
• Performance issues relating to the interaction of ISDN, B-ISDN, ATM, IP and cellular networks with the PSTN
• Management interfaces to speech enhancement equipment
• The control of speech enhancement devices by external devices
• The performance of automatic level control and speech enhancement devices in PSTN/GSTN ATM/IP and cellular networks
• The effects of residual acoustic echoes on the performance of the network echo canceller
• Hardware and software tools necessary to support the testing of speech enhancement devices
• Realistic echo paths and test signals required for testing speech enhancement devices
• Trade-offs between echo path delay, convergence quality, and convergence time of an echo canceller
• The impact of evolving terminal equipment such as modems, fax, and multimedia terminals on speech enhancement devices
• Performance issues relating to far-end echo cancelling and far-end delay build-out
• The simplification and repeatability of performance tests for speech enhancement devices and the ability of these tests to be independent of different manufacturers’ implementations and to represent real-world network operation
• The effect of clock differences between the receive and send paths of an echo canceller
• The effect of tandem coding avoidance (vocoder bypass or TFO) on in-path SPNE

Specific tasks

• Revise Recommendation G.168 by 2002 and Recommendation G.169 by 2004
• Develop draft new Recommendation(s) on equipment aspects of speech enhancement devices by 2002
• Develop draft new Recommendation on speech enhancement functions associated with wireless networks by 2002
• Develop draft new Recommendation on acoustic network echo controllers not associated with wireless networks by 2004
• Develop guidelines for the application of speech enhancement devices to be included in the respective Recommendations above

Relationships

• Other SG 15 questions as relevant
• ITU-T SG 4 on TMN aspects
• ITU-T SG 8 on facsimile
• ITU-T SG 11 on signalling issues
• ITU-T SG 12 on application guidelines and subjective performance
• ITU-T SG 13 on ATM and IP issues
• ITU-T SG 16 on multimedia issues
• ATM Forum and IETF
• ITU-R SG4 WP4B on voice on mixed terrestrial/satellite networks