Unequal error protection for perceptual audio coders

Details Unequal error protection for perceptual audio coders Unequal error protection for perceptual audio coders

1998-02-11

2002-06-11

Decady, Albert (Department: 2786)

Error detection/correction and fault detection/recovery

Pulse or data error handling

Digital data error correction

C375S245000, C375S253000, C704S500000, C704S501000, C704S502000, C704S503000, C704S504000, C714S755000

Reexamination Certificate

active

06405338

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates generally to audio coding devices, and more particularly to techniques for providing unequal error protection (UEP) for different classes of bits encoded in a perceptual audio coder (PAC) or other audio source coding device.
BACKGROUND OF THE INVENTION
Most source coded bit streams exhibit unequal sensitivity to bit errors. For example, certain source bits can be much more sensitive to transmission errors than others. Moreover, errors in certain bits, such as control bits, may lead to severe error propagation and a corresponding degradation in reconstructed signal quality. Such error propagation can occur, for example, in the output audio bits of an audio coder due to the use of control bits for codebook information, frame size information, synchronization information and so on. The perceptual audio coder (PAC) described in D. Sinha, J. D. Johnston, S. Dorward and S. R. Quackenbush, “The Perceptual Audio Coder,” in Digital Audio, Section 42, pp. 42-1 to 42-18, CRC Press, 1998, which is incorporated by reference herein, attempts to minimize the bit rate requirements for the storage and/or transmission of digital audio data by the application of sophisticated hearing models and signal processing techniques. In the absence of channel errors, a PAC is able to achieve near stereo compact disk (CD) audio quality at a rate of approximately 128 kbps. At a lower bit rate of 96 kbps, the resulting quality is still fairly close to that of CD audio for many important types of audio material.
The rate of 96 kbps is particularly attractive for FM band transmission applications such as in-band digital audio broadcasting (DAB) systems, which are also known as hybrid in-band on-channel (HIBOC), all-digital IBOC and in-band adjacent channel (IBAC)/in-band reserve channel (IBRC) DAB systems. There is also a similar effort underway to provide digital audio broadcasting at lower audio bit rates in the AM band. For these AM systems, audio bit rates of about 32 to 48 kbps are being considered for daytime transmission and about 16 kbps for nighttime transmission. Higher audio bit rates, greater than about 128 kbps, are being used in multiple channel DAB systems. The transmission channels in the above-noted DAB systems tend to be severely bandlimited and noise limited at the edge of a coverage area. For mobile receivers, fading is also a severe problem. It is therefore particularly important in these and other applications to design an error protection technique that is closely matched to the error sensitivity of the various bits in the compressed audio bit stream.
PACs and other audio coding devices incorporating similar compression techniques are inherently packet-oriented, i.e., audio information for a fixed interval (frame) of time is represented by a variable bit length packet. Each packet includes certain control information followed by a quantized spectral/subband description of the audio frame. For stereo signals, the packet may contain the spectral description of two or more audio channels separately or differentially, as a center channel and side channels (e.g., a left channel and a right channel). Different portions of a given packet can therefore exhibit varying sensitivity to transmission errors. For example, corrupted control information leads to loss of synchronization and possible propagation of errors. On the other hand, the spectral components contain certain interframe and/or interchannel redundancy which can be exploited in an error mitigation algorithm incorporated in a PAC codec. Even in the absence of such redundancy, the transmission errors in different audio components have varying perceptual implications. For example, loss of stereo separation is far less annoying to a listener than spectral distortion in the mid-frequency range in the center channel.
Unequal error protection (UEP) techniques are designed to match error protection capability with sensitivity to transmission errors, such that the most important bits are provided with the highest level of error protection, while less important bits are provided with a lesser level or levels of error protection. A conventional two-level UEP technique for use in DAB applications is described in N. S. Jayant and E. Y. Chen, “Audio Compression: Technology and Applications,” AT&T Technical Journal, pp. 23-34, Vol. 74, No. 2, March-April 1995. In this technique, which is based on a Reed-Solomon (RS) code, the control information is protected more robustly since it is not possible to use error mitigation on the non-redundant control information. In fact, the proper operation of the error mitigation algorithm used in a PAC codec is itself dependent upon reliable control information. All of the non-control spectral information in this technique is protected using a uniform level of error protection.
SUMMARY OF THE INVENTION
We have realized that a significant problem with the above-described unequal error protection (UEP) technique and other conventional UEP techniques is that these techniques fail to exploit the unequal impact on perceived audio quality of transmission errors in various spectral components. Moreover, the conventional techniques are often not directly applicable to a wide variety of bit rates and to transmission channels other than radio broadcasting channels.
The present invention provides methods and apparatus for implementing UEP in a source coded bit stream such as that generated by a perceptual audio coder (PAC). In an illustrative embodiment, an audio information bit stream includes audio control bits and audio data bits. The data bits are first separated into n different classes of bits, where n is greater than or equal to two, based on the sensitivity of the bits to source and channel errors. Each of the n different classes of data bits is then encoded in accordance with a corresponding one of n different levels of error protection. The invention thus separates audio data bits into classes based on the impact of errors in the data bits on perceived quality of an audio signal reconstructed from the transmitted encoded audio, and matches the error protection provided for the audio data bits to error sensitivity. For example, the data bits may be separated such that data bits corresponding to a side audio channel are assigned to a class which receives a lower level of error protection than data bits corresponding to a center audio channel. As another example, data bits corresponding to designated frequency range, such as 100 Hz to 4 kHz, may be assigned to a class which receives a higher level of error protection than data bits corresponding to another frequency range.
The audio control bits may be encoded independently of the audio data bits, using an additional level of error protection higher than that used for any of the n classes of the data bits. Alternatively, the control bits may be combined with one of the n classes of bits and encoded in accordance with the highest level of error protection of the n levels. Further protection may be provided for the control bits by repeating at least a portion of the control bits from a current packet in the audio bit stream in a subsequent packet of the audio bit stream. The UEP techniques of the invention may be designed to provide the same total bit rate as an equal error protection (EEP) technique.
The invention is applicable not only to PACs but also to other types of audio compression techniques operating over a wide range of bit rates, and can be used with transmission channels other than radio broadcasting. The above-described classification of data bits into n different classes can be implemented on a fixed packet-by-packet basis, or in a more flexible, adaptive implementation in which different multipacket error protection profiles are used for different multipacket segments of a source-coded audio signal. The invention can be implemented with a variety of different types of codes, including Reed-Solomon codes and rate-compatible punctured convolutional (RCPC) codes, as well as other types of block and convol

Affiliated with

Also associated with

Rating

Unequal error protection for perceptual audio coders does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Unequal error protection for perceptual audio coders, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Unequal error protection for perceptual audio coders will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2971454