Turbo Codes in AMC Systems For Blind Identification

Adaptive modulation and coding (AMC) systems are essen tial in Blind identification for channel codes. Sinc e Turbo codes are popular in AMC systems, it’s necessa ry to identify its parameters. In this paper, we foc us on the identification for Turbo codes from a closed-set. T he proposed approach firstly identifies the first c omponent code by accumulating Log-Likelihood Ratio (LLR) for syndrome a posteriori probability, then the interleaver and the other component code are identified by decoding based on zero insertion and LLR accumulation. This approach is robust to noise due to LLR. Moreover, it applies to both symme tric Turbo codes with two same component codes and a symmetric Turbo codes with two different component codes. Simu lation results demonstrate that the proposed blind identification scheme is able to identify Turbo codes at signal-to -noise ratio (SNR) larger than 3.5dB. The improvement for identification of RSC2 and the application for Turbo codes with various rates. we design a novel blind enc oder parameter estimator for turbo codes. In this scheme, we propos e to separate the feedback components from the forw ard path in a recursive convolution encoder so as to blindly esti mate the parameters. The simulations demonstrate th average estimation performance can lead to more than 95% ac curacy for the channel condition with the signal-to noise ratio at 5 dB. Once the encoder parameters are blindly estimate d, the corresponding decoder can be implemented to retrieve the original information sequences.