The paper presents a low complexity, low delay variable rate speech coding method suitable for low power (CDMA or TDMA) indoor wireless communications. The indoor wireless communications channel is characterized by relatively deep fades resulting in bursty errors. A robust speech codec is required to combat this bursty nature. Low power operation is essential, which negates the applicability of LD-CELP (G.728) for indoor wireless applications, Historically, ADPCM (G.721) has been proposed, which lacks robustness to random and bursty errors and has a relatively high bit rate (32 kb/s). To circumvent the need for relatively complex echo-cancellation, low end to end delays are required which implies that interleaving is not applicable. In the paper, a low complexity, low delay, variable rate CELP algorithm is proposed and techniques to improve the robustness to bursty errors are investigated. The new technique is compared to ADPCM on the basis of complexity and performance (MOS) over measured indoor wireless channels at 1.7 GHz with various combinations of space, frequency and code diversity.>

32 kb/s adaptive differential pulse coded modulation (ADPCM) is a widely accepted high quality speech coding algorithm. Single chip devices exist for 32 kb/s ADPCM. This algorithm requires a relatively large bandwidth and does not perform exceptionally well in noisy wireless environments. The 16 kb/s low-delay code excited linear prediction (LD-CELP) coding algorithm does perform well in noisy environments and requires a reduced bandwidth. Through algorithm optimization and simulated verification of mixed analog and digital VLSI partitioning a single chip implementing the 16 kb/s LD-CELP voice coding algorithm will be possible. The performance of this coder chip will be comparable to the quality of 32 kb/s ADPCM. In this paper we propose the use of a link between Comdisco's SPW and Synopsis to determine the hardware related performance degradation and the technology/complexity related power dissipation and area parameters. It will both prove the concept of design and determine the VLSI implementation feasibility.>

A tree-searched multistage vector quantization (VQ) scheme for linear prediction coding (LPC) parameters which achieves spectral distortion lower than 1 dB with low complexity and good robustness using rates as low as 22 b/frame is presented. The M-L search is used, and it is shown that it achieves performance close to that of the optimal search for a relatively small M. A joint codebook design strategy for multistage VQ which improves convergence speed and the VQ performance measures is presented. The best performance/complexity tradeoffs are obtained with relatively small size codebooks cascaded in a 3-6 stage configuration. It is shown experimentally that as the number of stages is increased above the optimal performance/complexity tradeoff, the quantizer robustness and outlier performance can be improved at the expense of a slight increase in rate. Results for log area ratio (LAR) and line spectral pairs (LSPs) parameters are presented. A training technique that reduces outliers at the expense of a slight average performance degradation is introduced. The method significantly outperforms the split codebook approach.>

The authors present a tree searched multi-stage vector quantization (TS-MSVQ) scheme which achieves spectral distortion lower than 1 dB with low complexity and good robustness using 24 b/frame. The M-L search is used and it is shown that it achieves performance close to that of the optimal search for a relatively small M. The best performance/complexity trade-offs are obtained with relatively small size codebooks cascaded in a three-four stage configuration. Results for log-area ratio (LAR) and line spectral pain (LSP) parameters are presented. A training technique which reduces outliers at the expense of a slight average performance degradation is introduced. The robustness across different languages and input spectral shapings is studied. Finally, it is shown that TS-MSVQ significantly outperforms the split-codebook approach.>