Browsing by Author "Lee, Cheng-Chieh"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Design of Structured Quantizers Based on Coset Codes
    (1995) Lee, Cheng-Chieh; Farvardin, N.; ISR
    For memoryless sources, Entropy-Constrained Scalar Quantizers (ECSQs) can perform closely to the Gish-Pierce bound at high rates. There exist two fixed-rate variations of ECSQ -- Scalar- Vector Quantizer (SVQ) and Adaptive Entropy-Coded Quantizer (AECQ) -- that also perform closely to the Gish-Pierce bound. These quantization schemes have approximately cubic quantization cells while high-rate quantization theory suggests that quantization cells of the optimal quantizers should be approximately spherical. There are some coset codes whose Voronoi regions are very spherical. In this dissertation we present structured quantization schemes that combine these coset codes with the aforementioned quantizers (SVQ, ECSQ, and AECQ) so as to improve their performance beyond the Gish-Pierce bound.

    By combining trellis codes (that achieve a significant granular gain) with SVQ, ECSQ, and AECQ, we obtain Trellis-Based Scalar- Vector Quantizer (TB-SVQ), Entropy-Constrained Trellis- Coded Quantizer (ECTCQ), and Pathwise-Adaptive ECTCQ (PA-ECTCQ), respectively. With an 8-state underlying trellis code, these trellis-coded quantization schemes perform about 1.0 dB better than their naive counterparts. There are two approaches that can extend the quantizers (TB-SVQ, ECTCQ, and PA-ECTCQ) for quantizing sources with memory. The first is to combine the predictive coding operation of the Differential Pulse Code Modulation scheme with various quantizers, yielding Predictive TB-SVQ, Predictive ECTCQ, and Predictive PA-ECTCQ, respectively. There is a duality between quantizing sources with memory and transmitting data over channels with memory. Laroia, Tretter, and Farvardin have recently introduced a precoding idea that helps transmitting data efficiently over channels with memory. By exploiting this duality, the second approach combines the precoder with TB-SVQ and ECTCQ to arrive at Precoded TB-SVQ and Precoded ECTCQ, respectively. Simulation results indicate that the porformance of these quantizers are also close to the rate- distortion limit.

    The PA-ECTCQ performance has been shown to be robust, in the presence of source scale and, to a lesser extent, shape mismatch conditions. We also considered adjusting the underlying entropy encoder based on the quantized output (which provide some approximate information on the source statistics). The performance of the resulting Shape-Adjusting PA-ECTCQ has been shown to be robust to a rather wide range of source shape mismatch conditions. are also close to the rate-distortion limit.

  • Thumbnail Image
    Item
    Entropy-Constrained Trellis Coded Quantization: Implementation and Adaptation
    (1993) Lee, Cheng-Chieh; Farvardin, Nariman; ISR
    Entropy-constrained trellis coded quantization (ECTCQ) of memoryless sources is known to be an efficient source coding technique in the rate-distortion sense. We develop an ECTCQ scheme that employs a symmetric reproduction codebook. The symmetry of the reproduction codebook, while essentially costs no performance loss, is exploited to reduce the memory requirement in entropy coding the ECTCQ output. In practice, a buffer of finite, and preferably small, size is needed to interface the variable-length codewords to the fixed-rate channel. An adaptive ECTCQ (A-ECTCQ) scheme, which uses a buffer-state feedback to control the quantizer characteristics to avoid buffer overflow/underflow, is studied in this work. The choice of encoding delay is an important issue in A- ECTCQ, as too long a delay will adversely impact the performance of the feedback control. We propose a pathwise-adaptive ECTCQ (PA-ECTCQ) that solves the encoding delay problem. Simulation results indicate that, while the buffer overflow/underflow problems of the PA- ECTCQ can be practically eliminated, the overall quantization distortion is increased only negligibly over theoretical performance predictions. Our experiments also suggests that PA- ECTCQ is robust with respect to source mismatch.