In this letter we present a low-complexity architecture designed for the decoding of block turbo codes. In particular we simplify the implementation of Pyndiah's algorithm by not storing any of the concurrent codewords generated by the list decoder.

This letter presents an original phase synchronization scheme designed for block turbo coded systems. This is a low-complexity scheme. It copes with low signal-to-noise ratios, taking complete advantage of the turbo coding possibilities. We display various results, showing that the proposed algorithm is able to process blocks of a few hundreds bits and that it tackles the case of large quadrature amplitude modulation constellations very well and with no phase ambiguity.

This paper is concerned with the design of block turbo codes for very high bit rate applications. We first introduce an original low-complexity architecture designed for the iterative decoding of product codes. The implementation of the Chase-Pyndiah algorithm is simplified by not memorizing any of the list decoding concurrent code words. We then illustrate that such block turbo codes allow some bit rate improvement in the context of local loop transmission

Dans ce travail, nous nous interessons au probleme du codage canal, dont le but est de corriger les erreurs dues au canal lors d'une transmission numerique. En particulier, les turbo codes constituent la derniere avancee dans ce domaine et atteignent la borne predite par C. E. Shannon en 1948. Nous presentons une simplification de la mise en oeuvre du turbo decodeur de R. Pyndiah, ainsi que des systemes de synchronisation profitant des informations apportees par le turbo decodeur

This paper concerns a new synchronizing scheme designed for block turbo coded system. The phase estimation relies on the feedback of the extrinsic information. A near optimum version of the algorithm based on the block turbo decoder devised by Pyndiah allows easy implementation. We display various results showing that the proposed algorithm is able to track frequency shifts and that it works very well in the case of general Quadrature Amplitude Modulation (QAM) constellations. We also show that this turbo synchronizer works at low Signal to Noise Ratios (SNR) taking completely advantage of the turbo coding possibilities.

The paper concerns a new synchronizing scheme designed for block turbo coded systems. Phase estimation relies on the feedback of reliability information from the output of the turbo decoder. It is based on the fact that the power of the lower reliability values is strongly dependent on the carrier phase offset. A near optimum version of the algorithm based on a Pyndiah-like block turbo decoder allows easy implementation. We display various results showing that the proposed algorithm is able to track frequency shifts and that it works very well in the case of general quadrature amplitude modulation (QAM) constellations. We also show that this turbo synchronizer works at low signal-to-noise ratios (SNR), taking complete advantage of the turbo coding possibilities.

We investigate the performance of different block turbo codes in the context of Enhanced VDSL. We first describe the VDSL system used in our simulation link. We then propose an iterative decoding algorithm based on Pyndiah's (1998) SISO decoder for product turbo block codes that can be implemented with reasonable complexity. We also illustrate several results for various QAM modulation schemes. The coding gains obtained for the different codes and modulations then allow us to predict the attainable bit rates as a function of the line length. We show that such a turbo block decoding outperforms the classical Reed Solomon hard decoding used in the ADSL system.

The phase estimation is a really important step in digital communication systems. Without a good synchronization, using error correction codes and turbo-codes is suboptimum. This paper describes a new synchronizing algorithm which can update the phase of each received symbol. The estimated phase is calculated with a Soft Decision Feedback Loop. This loop uses the Log- Likelihood Ratio (LLR) provided by the turbo-decoder's soft output. A forward-backward loop improves the phase estimation. This algorithm can be applied on every QAM's constellations and good performance is achieved till near Shannon's limit SNRs.

This paper presents a new synchronizing scheme designed for block turbo coded systems. The phase estimation relies on the feedback of the extrinsic information. A near optimum version of the algorithm based on the block turbo decoder devised by Pyndiah (1998) allows easy implementation. We display various results showing that the proposed algorithm is able to track frequency shifts and that it works very well in the case of general quadrature amplitude modulation (QAM) constellations. We also show that this turbo synchronizer works at low signal to noise ratios (SNR) taking completely advantage of the turbo coding possibilities.

This paper presents a new synchronizing scheme designed for block turbo coded system. The phase estimation relies on the feedback of the extrinsic information. A near optimum version of the algorithm based on the block turbo decoder devised by Pyndiah allows easy implementation. We display various results showing that the proposed algorithm is able to track frequency shifts and that it works very well in the case of general Quadrature Amplitude Modulation (QAM) constellations. We also show that this turbo synchronizer works at low Signal to Noise Ratios (SNR) taking completely advantage of the turbo coding possibilities. Résumé— : Ce papier présente une nouvelle technique de synchronisation de systèmes exploitant des turbo-codes blocs. L’estimation de la phase repose sur l’exploitation de la valeur extrinsèque en sortie du décodeur pour estimer cet écart en phase conjointement au décodage itératif des symboles. Nous présentons plusieurs résultats montrant que l’algorithme proposé permet de poursuivre des décalages en fréquence, qu’il fonctionne pour des modulations QAM et qu’il fonctionne à faible rapport signal à bruit (RSB). Une comparaison de la variance de cet estimateur à la borne de Cramer-Rao vraie (CRB) et modifiée (MCRB) est également effectuée.