Maja Loncar

Iterative channel estimation and data detection in frequency‐selective fading MIMO channels

Signals transmitted through multiple-input multiple-output (MIMO) wireless channels buffer from multiple-access interference (MAI), multipath propagation and additive noise. Iterative multiuser algorithms mitigate these signal impairments. while offering a good tradeoff between performance and complexity. The receiver presented in this paper performs channel estimation. multiuser detection and decoding in an iterative manner. The estimation of the frequeucy selective. block-fading channel is initiated with the pilot symbols. In subsequent iterations. soft decisions of all the data symbols are used in an appropriate way to improve the channel estimates. This approach leads, to significant Improvement of the overall receiver performance, compared to other schemes. The bit-error-rate (BER) performance of the receiver is evaluated by simulations for different parameter setups. Copyright (C) 2004 AEI. (Less)

Improved channel estimation for iterative receivers

In iterative receiver structures, soft information becomes available after the decoding stage. This information is used to enhance the quality of the channel estimates for the next iteration. We derive a generalized estimator based on the linear minimum mean square error (LMMSE) principle for deterministic pilot information combined with soft information. We present the special case of multi-carrier code division multiple access (MC-CDMA) in detail and provide simulation results. The presented channel estimation algorithm can be also applied to direct sequence (DS)-CDMA and multiple-input multiple-output (MIMO) systems.

BEAST decoding for block codes

BEAST is a Bidirectional Efficient Algorithm for Searching code Trees. In this paper, it is used for decoding block codes over a binary-input memoryless channel. If no constraints are imposed on the decoding complexity (in terms of the number of visited nodes during the search), BEAST performs maximum-likelihood (ML) decoding. At the cost of a negligible performance degradation, BEAST can be constrained to perform almost-ML decoding with significantly reduced complexity. The benchmark for the complexity assessment is the number of nodes visited by the Viterbi algorithm operating on the minimal trellis of the code. The decoding complexity depends on the trellis structure of a given code, which is illustrated by three different forms of the generator matrix for the (24, 12, 8) Golay code. Simulation results that assess the error-rate performance and the decoding complexity of BEAST are presented for two longer codes.

Low-complexity error correction of Hamming-code-based LDPC codes

Ensembles of binary random LDPC block codes constructed using Hamming codes as constituent codes are studied for communicating over the binary symmetric channel. These ensembles are known to contain codes that asymptotically almost meet the Gilbert-Varshamov bound. It is shown that in these ensembles there exist codes which can correct a number of errors that grows linearly with the code length, when decoded with a low-complexity iterative decoder, which requires a number of iterations that is a logarithmic function of the code length. The results are supported by numerical examples, for various choices of the code parameters.

A Comparison of Ungerboeck and Forney Models for Reduced-Complexity ISI Equalization

This paper investigates the performance of reduced- state trellis-based ISI equalizers, which are based on the so- called Ungerboeck and Forney observation models. Although the two models are equivalent when the Viterbi or BCJR equalizer is employed, their performances differ significantly when using reduced-complexity methods. It is demonstrated that practical equalizers operating on the Forney model outperform those operating on the Ungerboeck model for high signal-to-noise ratios (SNRs), while the situation is reversed for low SNR levels. A novel theoretical reduced-complexity equalization strategy that improves on previous Ungerboeck-based equalizers is proposed.

On Reduced-Complexity Equalization Based on Ungerboeck and Forney Observation Models

This correspondence investigates the asymptotic behavior of reduced-state intersymbol interference (ISI) equalizers, based on the Ungerboeck and Forney observation models. It is shown that Ungerboeck-based equalizers can suffer from correct-path-loss (CPL) even in the noiseless regime. ISI conditions that lead to CPL are analytically derived and illustrated by examples.

An Improved Bound on the List Error Probability and List Distance Properties

List decoding of binary block codes for the additive white Gaussian noise (AWGN) channel is considered. The output of a list decoder is a list of the most likely codewords, that is, the signal points closest to the received signal in the Euclidean-metric sense. A decoding error occurs when the transmitted codeword is not on this list. It is shown that the list error probability is fully described by the so-called list configuration matrix, which is the Gram matrix obtained from the signal vectors forming the list. The worst case list configuration matrix determines the minimum list distance of the code, which is a generalization of the minimum distance to the case of list decoding. Some properties of the list configuration matrix are studied and their connections to the list distance are established. These results are further exploited to obtain a new upper bound on the list error probability, which is tighter than the previously known bounds. This bound is derived by combining the techniques for obtaining the tangential union bound with an improved bound on the error probability for a given list. The results are illustrated by examples.

On channel estimators for iterative CDMA multiuser receivers in flat Rayleigh fading

In this work we compare the channel estimation algorithms for use in an iterative CDMA receiver in a block fading environment. The receiver consists of a soft multiuser data estimator, a bank of single user decoders, und a multiuser channel estimator. The multiuser data estimator is implemented as parallel interference canceler with unconditional post-MMSE filtering (PIC-MMSE) and the decoder is a soft-in soft-out MAP decoder. In the channel estimator we make use of dedicated pilot symbols and fed back soft-code symbols which are exploited as additional soft pilot symbols when the iterations proceed. We show that using extrinsic information increases the receiver performance significantly compared to using a posteriori information in the feedback for channel estimation. We introduce a linear MMSE (LMMSE) estimator which takes into account the variances of fed back code symbols and compare it to approximations of the least-squares (ALS) estimator and the linear minimum-mean-square-error (ALMMSE) estimator. Performance results are illustrated in terms of bit error rate (BER) and average normalized square error (ANSE) of the channel estimators. They show that the newly proposed LMMSE algorithm outperforms the ALS and ALMMSE algorithms.

Co-channel interference mitigation in GSM networks by iterative estimation of channel and data

In multiple-access communication systems several users share the system resources, thus creating co-channel interference to each other. Multiuser detection, applied at the receiver side, is a powerful technique to combat co-channel interference, in contrast to single-user approaches that treat the interference as additive noise. In this paper, we investigate the feasibility of achieving increased spectral efficiency in GSM-like TDMA networks by applying multiuser detection. We focus on downlink operation and consider two mobile users that share the same physical channel within the same cell or in two adjacent cells. No synchronization between the base stations is assumed. Mobile terminals exploit multiple receive antennas. We propose a simple iterative receiver algorithm that performs joint channel and data estimation. Results of simulations that evaluate the receiver’s performance are presented. (Less)

On the asymptotic performance of low-complexity decoded LDPC codes with constituent hamming codes

Hamming code-based LDPC (H-LDPC) block codes are obtained by replacing single parity-check codes in Gallagerpsilas LDPC codes with Hamming constituent codes. This paper investigates the asymptotic error-correcting capabilities of ensembles of random H-LDPC codes, used over the binary symmetric channel and decoded with a low-complexity hard-decision iterative decoding algorithm. The number of required decoding iterations is a logarithmic function of the code length. It is shown that there exist H-LDPC codes for which such an iterative decoding corrects any error pattern with a number of errors that grows linearly with the code length. For various choices of code parameters the results are supported by numerical examples.