Title :
Linear subcodes of turbo codes with improved distance spectra
Author :
Daneshgaran, Fred ; Mondin, Marina ; Mulassano, Paolo
Author_Institution :
Electr. & Comput. Eng. Dept., Calif. State Univ., Los Angeles, CA, USA
Abstract :
In this correspondence, we present a technique for generation of linear subcodes of a given turbo code with better distance spectrum than the original mother turbo code, via an iterative process of trace-bit injection which minimally reduces code rate, followed by selective puncturing that allows recovery of the rate loss incurred during the trace-bit injection. The technique allows for asymptotic performance improvement of any linear turbo code. In effect, we trim the distance spectrum of a turbo code via elimination of the low distance and/or high multiplicity codewords from the output space of the code. To this end, we perform a greedy minimization of a cost function closely related to the asymptotic bit error probability (or frame error probability) of the code. This improves the performance of the code everywhere, but its main impact is a reduction in the error floor of the turbo code which is important for delay constrained applications employing short interleavers.
Keywords :
error statistics; iterative methods; linear codes; minimisation; turbo codes; asymptotic bit error probability; code rate; cost function; delay constrained turbo codes; frame error probability; greedy minimization; improved distance spectra; iterative process; linear subcodes; selective puncturing; trace-bit injection; Concatenated codes; Convolutional codes; Cost function; Delay; Doping; Error probability; Floors; Iterative decoding; Performance gain; Turbo codes; 65; Code doping; delay constrained turbo codes; distance spectrum; interleavers; optimization; puncturing; short length turbo codes; subcode; trace-bit injection; turbo codes;
Journal_Title :
Information Theory, IEEE Transactions on
DOI :
10.1109/TIT.2004.838364