Gilles Mesnager

Interleaver design for turbo codes from convergence analysis

In this letter, we present an analysis tool for the convergence of a turbo decoder using the min-sum algorithm. First of all, we classify cycles on a turbo code factor graph. We define and develop a quantity which characterizes the trellis convergence and a probability of a message round on a cycle. Our analysis is then applied to an interleaver design for a turbo decoder using the min-sum algorithm. Since the probability of a message round is closely related to the convergence property, the constructed interleaver optimizes the convergence property. The performance is compared with several referential interleavers.

Modeling the burst error statistics of Viterbi decoding for punctured codes

One of the characteristics of the Viterbi algorithm is that errors at the output appear grouped in bursts. The knowledge of these error burst statistics is very useful in designing a communication system. This paper presents a simple model of Viterbi decoder burst-error statistics derived from the code distance properties. The puncturing process is also modeled. Results are validated through additional computer simulations. The advantage of the model is that it may be used to generate Viterbi error sequences quickly when a very large amount of data is needed. The complete analysis of the convolutional code used in Digital Video Broadcasting (DVB) standards is shown.

Interleaver design for serial concatenated convolutional codes

In this paper, we present an interleaver design method for SCCC. Our design criterion is to minimize the message-round probability in the SCCC graph which is specially well suited for SCCC. The message-round probability characterizes the message flow in the SCCC graph. By minimizing it, we can get a large interleaving gain. The simulation results confirm our approach.

Trade-off between spectral efficiency increase and PAPR reduction when using FTN signaling: Impact of non linearities

Faster-than-Nyquist (FTN) signaling appears as an attractive method to improve spectral efficiency at the price of an increased complexity at the receiver. The receiver generally implements a turbo-equalization/detection scheme to benefit from all the promises of the FTN signaling. However, this is not the only limitation we have to deal with. Indeed, compressing in the time domain impact the emitted signal and it usually results in an increase of the envelope fluctuations. This leads to an inherent multi-objectives trade-off between performance, targeted spectral efficiency and limited Peak to Average Power Ratio (PAPR). The last aspect is crucial when considering a satellite communication link due to non-linear amplification effects that can occur on-board the satellite. Usually, FTN studies focus on spectral efficiency increase for a fixed modulation order, trying to trade-off between performance and PAPR properties. In this paper, we show that, for a given asymptotic spectral efficiency, we can compress low order modulations to increase the spectral efficiency of these schemes while controlling the PAPR increase to achieve a better PAPR than the non compressed scheme with a higher modulation order. Thus, for the same asymptotic spectral efficiency, we can achieve 1 dB gain in terms of PAPR and 2 dB gain in Bit Error Rate (BER) performances for a coded 8-PSK FTN system compared to a coded 16-APSK. For the same BER performances, the asymptotic spectral efficiency gain obtained in linear context is over 20 %, higher when non-linearities are taken into account.

A soft detection directed phase estimator suited to satellite burst transmissions

In this paper, we present a phase synchronization algorithm which exploits the data redundancy brought by a single parity-check code. The symbol decoding of a parity equation is discussed first. Then, a soft detection directed (SDD) estimation technique is derived for any modulation scheme. The performance of the developed SDD phase estimator is evaluated with 8PSK modulation in very low SNR condition. It appears to greatly outperform the non data aided Viterbi and Viterbi estimator. Quite robust to phase noise, it seems well suited to the context of satellite burst transmissions

Error distribution in turbo-decoding with max-log-MAP algorithm

Since the apparition of turbocodes in 1993, this topic has been a great success and the subject of much research. But the performance of these coding systems are often presented only in terms of bit-error-rate (BER) or frame-error-rate (FER). These parameters are not the only ones to take into account when designing a communication system. Other properties of the error distribution are also important when analysing the performance of the system at higher levels such as the networking layer. This paper presents some properties of the errors at the output of a turbo decoding system. It also presents a model that is able to describe the error burst length when using the max-log-MAP decoding algorithm.

Modeling Burst Error Properties in Turbocodes: Application to DVB-RCS

One of the most studied fields in coding are turbocodes. Everyone knows the good performance of turbocodes in terms of bit error rate (BER). However, very little is known about the error distribution at its output. The knowledge of these error statistics is very useful to design the packet size or the overhead location. Are these errors randomly distributed within the packet, or do they follow a particular law? People working at physical layer level need more information about the error distribution at the output of the turbodecoder. This paper gives some statistics of the errors after turbodecoding. A model describing this error behavior is presented and validated through computer simulations for the turbocode used in the interaction channel for Digital Video Broadcasting by Satellite (DVB-RCS standard).