Summary

ARD0R objectives

The objective of the ARDOR (Adaptive Rate-Distortion Optimised sound codeR) project is to meet the need for a universal codec that has been created by the emergence of time-varying heterogeneous networks and by the convergence of traditional consumer electronics with mobile communications. Traditionally, codecs have been developed with a particular application in mind. The performance of such codecs strongly depends on the signal (speech or music) and the constraints associated with the application, e.g., bit rate. As a result, many standardised codecs exist and this number is increasing rapidly. Such a proliferation of codecs is undesirable since it means that the user or system designer is burdened with the task of selecting the right codec for a particular application. More importantly, with the emergence of time-varying heterogeneous network environments (fourth generation ubiquitous networks), any selection of a codec with a time-invariant configuration is most likely to be non-optimal. Universal codecs are a necessity in future ambient intelligent landscapes.

Figure 1. Schematic illustration of a ubiquitous network. The grey cloud visualises the heterogeneous network, hosting a variety of applications, such as Internet radio, (mobile) telephony, (portable) audio players and set-top boxes. The grey arrows (with various shapes) indicate the different codecs currently in use. The black/red arrows (all of equal shape) indicate the universal ARDOR codec

It is no coincidence that, when system complexity is growing rapidly in communication technology, there is much discussion about network appliances. There is a strong need for intelligent subsystems with a simple 'plug and play' property combined with high efficiency. A universal sound codec would satisfy these simultaneous demands for simplicity and efficiency when coding sound signals. This type of codec will result in the much-needed transparency to the user and eliminate the commonly observed mismatch between codec and application, particularly in the heterogeneous networks, as illustrated in Figure 1, that are predicted to become commonplace in the near future. For these reasons, we expect that universal codecs will become an integral part of the ambient intelligent landscape that is developing in society.

The ARDOR project is developing universal sound codec. This technology is able to adapt to the time-varying characteristics of the input signal, to user preferences, and to application-imposed constraints or time-varying network-imposed constraints on coding attributes such as bit rate or quality. As a result, this technology is useful in a broad range of applications, such as (video) telephony, Internet radio, (portable) audio players, conferencing systems, telephone answering devices, mobile communication, digital audio broadcast, solid-state audio playback systems, and set-top boxes.

ARD0R Approach

Figure 2. A schematic diagram of a rate-distortion optimised coder

The development of a truly adaptive codec is a challenge to the merging fields of speech and audio coding that require innovations on several fronts. Within the project, a codec is created that encodes generic audio with the most appropriate combination of signal models, given the imposed constraints as well as the available sub-coders. The work can be divided into three parts corresponding to the three codec components (see Figure 2): 

• a rate-distortion-theory based optimisation mechanism that configures the ARDOR codec such that it operates most efficiently given the current, time-varying, constraints and type of input signal. For this purpose it controls:

• a set of 'subcoding' strategies, each of which is highly efficient for encoding a particular type of input-signal component, e.g., tonal, noisy, or transient signals. The appropriate rate and signal-component allocation for each particular subcoding strategy is based on:

• an advanced, new perceptual distortion measure that provides a perceptual criterion for the rate-distortion optimisation mechanism. In other words, a perceptual model, which is based on state-of-the-art knowledge about the human auditory system, provides the optimisation mechanism with information about the perceptual relevance of different parts of the sound. The optimisation algorithm could for example decide to leave out information that is perceptually irrelevant. Consequently, the original signal cannot be restored, but the auditory system will not be able to perceive the difference.

These three codec components are integrated in a dedicated flexible software platform.

Expected results

The end result of the ARDOR project is a prototype of a universal sound codec. This codec is self-optimising, adaptive to time-varying constraints and user preferences, and it provides graceful degradation for decreased data rates.

 comment to: Project Office  .
last updated  

   copyright © {ARDOR, IST-2001-34095}