Claudio J. Bordin

Particle Filters for Joint Blind Equalization and Decoding in Frequency-Selective Channels

This paper introduces new algorithms for joint blind equalization and decoding of convolutionally coded communication systems operating on frequency-selective channels. The proposed method is based on particle filters (PF), recursively approximating maximum a posteriori (MAP) estimates of the transmitted data without explicitly determining channel parameters. Further elaborating on previous works, we assume that both the channel order and the noise variance are unknown random variables, and develop a new formulation for PF weight propagation which allows these quantities to be analytically integrated out. We verify via numerical simulations that the proposed methods lead to near optimal performance, closely approximating that of algorithms that require exact knowledge of all channel parameters.

Cooperative blind equalization of frequency-selective channels in sensor networks using decentralized particle filtering

We introduce in this paper new decentralized particle filtering algorithms suitable for blind equalization of frequency-selective communication channels. The proposed methods rely on novel distributed sequential importance sampling techniques that spread the computational load across a network of processing nodes, which cooperate in turn to produce a global consensus estimate of the transmitted data stream. As we verify via numerical simulations, the new decentralized schemes approach the performance of the optimal centralized MAP receiver, exhibiting clear performance improvements compared to previous data-blind methods.

Bayesian blind equalization of time-varying frequency-selective channels subject to unknown variance noise

We present in this article a novel particle-filter-based blind equalization algorithm suitable for FIR time-varying frequency-selective communication channels corrupted by unknown variance additive Gaussian noise. The proposed method is fully Bayesian, integrating out the unknown parameters via an original recursive method, unlike previous approaches that rely on suboptimal plug-in estimates. We verify via numerical simulations that the proposed methods performance approaches that of the trained MAP equalizer, exceeding that of the linear least squares Kalman equalizer for medium to low noise levels.

A new minimum-consensus distributed particle filter for blind equalization in receiver networks

We describe in this paper a novel distributed particle filtering algorithm that performs blind equalization of frequency-selective channels in a setup with a single transmitter and multiple receivers. The algorithm employs parallel minimum consensus iterations to determine some a posteriori probability functions, providing equal approximations on all network nodes in a finite, deterministic, network-dependent number of steps. We verify via computer simulations that the new algorithm exhibits a bit error rate (BER) performance similar to that of the centralized particle-filter estimator with communication requirements milder than that of previous approaches, as the new method drops the need to evaluate quantities via average consensus.

A Rao-Blackwellized Particle Filter for Blind Equalization of Frequency-Selective Channels with Unknown Order and Noise Variance

We propose in this paper a new particle filtering algorithm for blind equalization of FIR frequency-selective communication channels corrupted by additive Gaussian noise, assuming that both the channel order and noise variance are unknown. The proposed algorithm integrates out analytically the unknown parameters using a modified sequential importance sampling technique. We verify via numerical simulations that the proposed method leads to near optimal performance, greatly outperforming traditional methods under noise variance mismatch.

Particle filter algorithms for joint blind equalization/decoding of convolutionally coded signals

This work introduces the use of particle filters for joint blind equalization/decoding of convolutionally coded signals transmitted over frequency selective channels. As in the equalization-only case, we show how to evaluate the optimal importance function recursively via a bank of Kalman filters. Numerical simulation investigations using both stochastic and deterministic particle selection strategies show the outstanding superiority of the deterministic joint equalization/decoding method over approaches that perform blind equalization using particle filters prior to optimal decoding.

Deterministic particle filters for joint blind equalization and decoding on frequency selective channels

This work proposes deterministic particle filtering structures for joint blindly equalizing/decoding convolutionally coded signals transmitted over frequency selective channels. After describing the proposed structures, we show how to evaluate the weight update functions corresponding to the adopted signal model. Numerical simulations show that the algorithm employing deterministic particle selection greatly outperforms alternative stochastic strategies, even when the latter employ the optimal importance function

Joint blind equalization and decoding using particle filters

This work examines how particle filtering algorithms-numerical techniques for solving Bayesian estimation problems-can be used for joint blind equalization and decoding of block-coded signals transmitted over frequency selective channels. Via simulations, we verify that the performance of the proposed method is much better than that of an equalization-only algorithm, outperforming the optimal linear equalizer at moderate noise levels at almost no additional computational cost.

Particle Filters for Blind FIR Channel Equalization in Non-Gaussian Noise

In this work, we propose new particle filter based blind equalization algorithms for FIR channels subject to additive noise of arbitrary distributions. These algorithms employ artificial evolution methods to jointly generate samples from the missing data and from the unknown channel parameters, which are assumed time-invariant. To achieve these results we introduce a new importance function, which leads to greatly improved performance compared to more obvious alternatives as verified via numerical simulations using Weibull envelope noise processes, in which the performance of the trained MLSE equalizer is approached to a narrow margin

A new sequential Monte Carlo algorithm for distributed blind equalization

This article describes a new distributed blind equalization algorithm based on particle filters suitable for CDMA systems operating on frequency-selective channels. The proposed approach processes the signal received by multiple receivers in a non-centralized fashion, determining approximately optimal estimates (MAP) of the transmitted symbols, which minimizes the expected error rate. The described receiver structure dispenses with the need of a data fusion center, and is capable of operating either independently or in cooperation with other receivers. As we verified via numerical simulations, cooperative operation leads to a significant performance improvement, without substantially increasing the algorithmpsilas computational complexity.