List-Decoding of MPEG coded audio signals
R. Bauer, Harald Ernst · elib (German Aerospace Center) · 2000
| This paper proposes the application of a list output Viterbi algorithm (LOVA) to compressed high-quality audio data when either MPEG-1 layer 3 or MPEG-2AAC (Advanced Audio Coding) is used for data compression. In both cases, LOVA-decoding is shown to yield significant performance gains in terms of the resulting audio quality compared to conventional Viterbi decoding. Especially when error events can lead to error propagation effects during source decoding, the scheme is very effective. For both, the additive white Gaussian noise (AWGN) and the fully interleaved Rayleigh fading channel simulation results with the segmental SNR as measure for the achieved audio quality are shown. Finally we give a brief comparison of both standards in terms of error resilience and suitability for transmission over a mobile radio channel. I. INTRODUCTION In state of the art audio compression schemes sophisticated tools are used to reduce irrelevant and redundant parts of the original audio signal. In ...