Title :
Joint optimization of TCQ-TCM systems
Author :
Aksu, H.A. ; Salehi, Masoud
Author_Institution :
Dept. of Electr. & Comput. Eng., Northeastern Univ., Boston, MA, USA
fDate :
5/1/1996 12:00:00 AM
Abstract :
Signal set expansion and set partitioning ideas from trellis coded modulation (TCM) have been previously applied to develop a source coding technique known as trellis coded quantization (TCQ). It has been also known that joint optimization of source and channel encoders/decoders results in substantial performance improvement. In this work, we investigate the performance of joint TCQ/TCM systems and develop various system optimization techniques. We consider channel optimized TCQ levels and asymmetric TCM symbol sets, and propose a simulated annealing based algorithm for joint optimization of TCQ levels and TCM symbols. Using these techniques signal to quantization noise ratio (SQNR) gains of 0.5-4 dB over nonoptimized systems are attained off-line. This, in turn, provides all the gain of going to a one step higher order trellis
Keywords :
decoding; simulated annealing; source coding; trellis coded modulation; TCQ-TCM systems; asymmetric TCM symbol sets; channel encoders; decoders; higher order trellis; joint optimization; performance; set partitioning; signal set expansion; signal to quantization noise ratio; simulated annealing based algorithm; source coding technique; trellis coded modulation; trellis coded quantization; Convolution; Convolutional codes; Decoding; Degradation; Iterative algorithms; Modulation coding; Quantization; Signal to noise ratio; Simulated annealing; Source coding;
Journal_Title :
Communications, IEEE Transactions on