One of the essential components of any current communication system is the encoding for protection against disturbances. The exponentially increasing volume of data transmitted imposes the improvement of communication systems and some encoding and decoding methods against errors. The appearance of turbo-codes (TC), whose performance BER/SNR (Bit Error Rate /Signal to Noise Ratio), in terms of the coding gain, comes almost below 1 dB at the theoretical limit, has opened new research horizons.
This project proposes to continue the study of a family of turbo-codes, named multi-non-binary turbo-codes (MNBTC), which has been proposed for the first time by the Project Leader.
This research has the purpose to improve some details that characterize the construction and operation of the MNBTCs: to reduce the complexity and the volume of calculation for the turbo-decoder, to combat the phenomenon of "error floor", to minimize the delay caused by the turbo-decoding, to increase TCs adaptation to different types of channels (for example fading channels, special channels) or different coding rates, etc. ;
The proposed findings will be extended to the family of turbo codes that have Reed-Solomon codes as component codes. Also, another objective of this study is to extend the scope of using TCs in various applications (watermarking, optical communications, etc).