Valery Ramon

EM algorithm-based timing synchronization in turbo receivers

The paper addresses the issue of estimating the sampling instant in turbo receivers. The proposed synchronizer is based on the expectation-maximization (EM) algorithm and takes benefit from the soft information delivered by the turbo system. Performance of the proposed synchronizer is illustrated by simulation results. In particular, the mean and the variance of the estimator as well as the bit error rate reached by the synchronized system are reported.

A Semi-Analytical Method for Predicting the Performance and Convergence Behavior of a Multiuser Turbo-Equalizer/Demapper

This paper proposes a simple semi-analytical method with reduced simulation time for predicting at any iteration the performance of a turbo-equalization/demapping scheme using the Wang and Poor soft-in/soft-out (SISO) minimum mean-square error (MMSE)/interference cancellation (IC) equalizer and a SISO decoder. The proposed method may be applied to multilevel/phase data modulations as well as multiuser context. This paper shows that the equalizer behavior may be very reliably predicted totally by calculations (no simulations are needed) whereas that of the decoder still requires simulations. A comparison between the proposed prediction method and plain simulations of the overall turbo equalization scheme demonstrates that our method accurately determines the system performance at any iteration. Static channels, long frames, and perfect channel knowledge are assumed throughout the paper

Turbo synchronization: a combined sum-product and expectation-maximization algorithm approach

This paper places turbo synchronization into the sum-product (SP) and the expectation-maximization (EM) algorithm framework. In particular, we show that the combination of these algorithms enables to design low-complexity and very powerful synchronisers. The proposed synchronizer is compared to a previously-proposed EM framework. As the derivation suggests, our new iterative scheme clearly outperforms the classical use of the EM algorithm.

EM algorithm-based multiuser synchronization in turbo receivers

The current paper addresses the issue of estimating the user propagation delays, received carrier phase offsets and received amplitudes in an asynchronous DS-CDMA environment with frequency non-selective propagation channels. The proposed synchronizer is based on the expectation-maximization (EM) algorithm and takes benefit from the soft information delivered by the receiver which is of the turbo type. The performance of the proposed synchronizer is illustrated by simulation results. In particular, the mean and the mean squared error of the estimator, as well as the bit error rate reached by the synchronized system, are reported.

EM algorithm-based estimation of amplitude, carrier phase and noise variance in multiuser turbo receivers

The current paper addresses the issue of estimating the users received carrier phase offsets and received amplitudes as well as the noise variance in an asynchronous DS-CDMA environment with frequency nonselective propagation channels. The user propagation delays are assumed to be known or estimated beforehand. The proposed synchronizer is based on the expectation-maximization (EM) algorithm and takes benefit from the soft information delivered by the receiver which is of turbo type. Besides, this paper explains which training symbols should be preferably chosen to initialize this synchronizer. It also compares two methods for combining the training-based estimate and the EM estimate. Performance of the resulting synchronizer is illustrated by simulation results.

An iterative soft decision directed linear timing estimator

The paper addresses the issue of estimating the symbol timing in turbo receivers. We propose a linear timing estimator which benefits from the soft information delivered by the turbo system at each iteration to compute the timing estimate. This estimator is then compared to a synchronizer based on the expectation-maximization (EM) algorithm (Herzet, C. et al., ICASSP, 2003). Performance of the proposed synchronizer is illustrated by simulation results and proves better than that of the EM-based counterpart regarding both convergence speed and range.

Soft Estimation of Time-Varying Frequency Selective Channels Using Kalman Smoothing

This paper addresses soft estimation of time-varying frequency selective channels using Kalman smoothing. The proposed estimator uses soft extrinsic information provided by a channel decoder. It is intended to improve the performance of an already existing Kalman filtering-based estimator by exploiting all - rather than part of - the data at the receiver disposal. It is the linear estimator exhibiting for the case of interest the minimum mean-squared estimation error. Its complexity is shown to be quite low. An approximated analytical calculation of the mean- squared estimation error (MSEE), both for Kalman filtering and Kalman smoothing, is also proposed. Simulation results illustrate the performance gain of smoothing over filtering and validate our calculation of the MSEE.

Frame-Error-Rate-wise Optimal Code-Aided Hypothesis Testing

This paper addresses the issue of code-aided hypothesis testing in communication systems, i.e., the estimation of discrete-valued nuisance parameters. A hypothesis testing procedure optimal in the sense of the minimization of the frame-error rate (FER) is derived. Its complexity is shown to be comparable to other recently-proposed code-aided hypothesis procedures. Moreover, when a conditional maximum a posteriori decision rule is used to make the decisions about the transmitted sequence, it is shown that the proposed hypothesis testing procedure combined with sequence detection reduces, in a good approximation, to a joint maximum-likelihood problem. Finally, as an illustrative example, we show the case of phase ambiguity resolution for a convolutionally-coded transmission.

Predicting the performance and convergence behavior of a turbo-equalization scheme

This paper proposes a simple and time non-consuming method to predict, at any iteration, the performance of a turbo equalization scheme using a soft-in/soft-out (SISO) minimum mean square error (MMSE)/interference cancellation (IC) equalizer and a SISO decoder. Gaussianity of the extrinsic log-likelihood ratios (LLRs) output by the equalizer as well as the decoder is assumed. This paper shows that the equalizer behavior may then be very reliably predicted only by calculations (no simulations are needed) whereas that of the decoder requires simulations for only one independent input parameter. Comparison between the proposed prediction method and plain simulations of the overall turbo equalization scheme shows that our method accurately determines the system performance at any iteration.

Iterative Synchronization: EM Algorithm Versus Newton-Raphson Method

This paper deals with iterative maximum-likelihood synchronization of a scalar parameter. An efficient implementation of the Newton-Raphson (NR) maximum-search method is proposed. Considering the latter implementation, the NR approach is shown to be an attractive alternative to synchronization methods based on the expectation-maximization (EM) algorithm. Simulation results for the case of phase-offset synchronization show that NR method usually increases the speed of convergence of the synchronization algorithm