Multistage trellis coded quantisation (MS-TCQ) is developed as a constrained trellis source coding technique for successive refinement of information. The authors study the performance of two-stage trellis coded quantisers for Gaussian and uniform memoryless sources. Problems of appropriate trellis selection, output alphabet selection, and stage-by-stage design of the codebooks are addressed. Stagewise optimisation is done using the generalised Lloyd algorithm. Comparisons with two-dimensional multistage vector quantisation (MSVQ) indicate that up to 1.6 dB performance improvement can be achieved by using MS-TCQ. Another algorithm is also presented for the joint design of the stage codebooks to optimise the overall performance. It is shown that another 0.5 dB improvement can be achieved by employing the joint design algorithm.
[1]
Allen Gersho,et al.
Vector quantization and signal compression
,
1991,
The Kluwer international series in engineering and computer science.
[2]
William Equitz,et al.
Successive refinement of information
,
1991,
IEEE Trans. Inf. Theory.
[3]
Christopher F. Barnes,et al.
Residual vector quantizers with jointly optimized code books
,
1992
.
[4]
Lee-Fang Wei,et al.
Coded modulation with unequal error protection
,
1993,
IEEE Trans. Commun..
[5]
Michael W. Marcellin,et al.
Trellis coded quantization of memoryless and Gauss-Markov sources
,
1990,
IEEE Trans. Commun..