Soft-In Soft-Out Detection Using Partial Gaussian Approximation

Turbo codes are applied to magnetic recoding channels by treating the channel as a rate-one convolutional code that requires a soft a posteriori probability (APP) detector for channel inputs. The complexity of conventional APP detectors, such as the BCJR algorithm or the soft-output Viterbi algorithm (SOVA), grows exponentially with the channel memory length. This paper derives a new APP module for binary intersymbol interference (ISI) channels based on minimum mean squared error (MMSE) decision-aided equalization (DAE), whose complexity grows linearly with the channel memory length, and it shows that the MMSE DAE is also optimal by the maximum a posteriori probability (MAP) criterion. The performance of the DAE is analyzed, and an implementable turbo-DAE structure is proposed. The reduction of channel APP detection complexity reaches 95% for a five-tap ISI channel when the DAE is applied. Simulations performed on partial response channels show close to optimum performance for this turbo-DAE structure. Error propagation of the DAE is also studied, and two fixed-delay solutions are proposed based on combining the DAE with the BCJR algorithm.

The paper presents a novel fast-adaptive nonlinear receiver which exploits soft statistics for tracking the random fluctuations experienced by time division multiple access (TDMA) mobile radio links impaired by frequency-selective time-variant multipath phenomena. The detection task is accomplished by an Abend-Fritchman-like symbol-by-symbol maximum likelihood (SbS-ML) detector which delivers both hard decisions and soft statistics in form of a posteriori probabilities (APPs) of the states of the intersymbol interference (ISI) channel. In the proposed adaptive receiver, these APPs are employed in place of the conventional hard-detected data to feed an ad hoc developed nonlinear recursive Kalman-type channel estimator. Extensive computer simulations show that the exploitation of soft statistics enhances the tracking capability of the channel estimator so that the proposed receiver generally outperforms the usual ones based on adaptive maximum likelihood sequence estimators (MLSEs) for signal-to-noise ratio (SNR) values over 12-13 dB. Furthermore, the experienced performance gap with respect to more complex per-survivor processing (PSP)-based multi-estimator detectors appears generally small on slowly and moderately fast time-varying channels characterized by values of the product Doppler bandwidth /spl times/ signaling period B/sub D/T/sub S/ below 5/spl times/10/sup -3/.

The subject of turbo coding in the presence of an intersymbol interference (ISI) channel whose characteristics are unknown is investigated. An iterative blind turbo detector, is presented, which combines the channel estimation, the channel equalization, and the turbo decoding. Simulation results are presented for rate 1/2 recursive systematic convolutional (RSC) codes and rate 1/3 turbo codes with BPSK modulation, transmitted over an ISI channel having severe frequency distortion. We can improve the performance when we use a good choice of RSC encoder of which the denominator is a primitive polynomial.

Summary form only given. Following the work done by Ryan (Proc. ICC Int. Conf. on Communications, 1998), considerable research has been reported on the application of concatenated codes to partial response (PR) channels. Both parallel concatenated codes (PCC) and serial concatenated codes (SCC) have been studied. In all of those works, a linear model for the channel was assumed, which is a valid assumption for low recording densities. However, as the recording density increases, the channel becomes nonlinear where nonlinearity arises from closely spaced recorded transitions. In this paper, we combine concatenated coding and nonlinear equalization for PR channels that suffer from nonlinear distortion (or intersymbol interference) according to the setup shown in Fig 1. We extend the RAM-search/Viterbi-detector concept to the a posteriori probability (APP) and soft-output Viterbi algorithm (SOVA) detectors that are used with turbo-coded partial response channels. The modified APP and SOVA are called RS-APP and RS-SOVA, respectively.

In this paper, we present a semi‐analytical approach for analyzing the serial concatenation (SCC) of a high rate convolutional code and a precoded intersymbol interference (ISI) channel. The significance of the proposed approach is that it can be used to identify, for a given outer high rate code and fixed ISI channel, the precoder that results in the lowest error rate floor. In estimating the bit error performance in the floor region, we use analytical techniques developed for trellis codes to compute the overall system minimum squared Euclidean distance, and then use a semi‐analytical approach to find the corresponding multiplicity. We also demonstrate via simulations that the proposed technique may be extended to fading ISI channels with favorable results. We give examples that support our analysis. Copyright © 2006 John Wiley & Sons, Ltd.

The authors consider the use of multidimensional signaling and block demodulation for digital transmission over intersymbol interference (ISI) channels. As an example, the theory is developed for the unit-memory 1+ alpha z/sup -1/ channel. The authors first explain why multidimensional signals should be used over ISI channels by showing how the minimum Euclidean distance and the cutoff rate R/sub 0/ increase by increasing the dimensionality of the signals transmitted. Two ways of increasing the reliability of transmission over an ISI channel are discussed. The first way consists of increasing the energy associated with the symbols more affected by ISI. The second way uses a coding scheme which selectively protects those symbols. Generalized concatenated-codes are advocated for this purpose. >

Wiener filter is widely used to equalize the signals over inter-symbol interference (ISI) channels because of its low complexity and optimality in terms of mean square error (MSE). However, Wiener filter is generally not optimal in terms of symbol detection error and thus can exhibit considerable error performance degradation in ISI channels. Recently, to enhance the performance of linear equalizers, lattice reduction (LR) technique is proposed, which shows significant gain for multi-input multi-output systems. In this paper, we propose LR-aided Wiener filter that aims at minimizing the MSE for the symbols in lattice-reduced domain. Simulations show that the LR-aided Wiener filter exhibits significant error performance improvement over the Wiener filter, especially for higher-order modulations.

We present a novel iterative (turbo) receiver with tunable complexity for reliable detection of (uncoded) payload data transmitted over long intersymbol interference (ISI) channels affected by crosstalk, as those typically encountered in emerging HDSL2 services standardized by ANSI T1.418 recommendation. The proposed receiver combines in an original way "minimum mean square error (MMSE) estimation principle," "turbo-processing principle" and "crosstalk-prediction principle" for achieving both suboptimal maximum a posteriori probability channel equalization and reliable soft-mitigation of ISI tail plus crosstalk. More in detail, according to the turbo-processing principle, at each iteration suitable extrinsic information is extracted from the equalizing and interference-canceling modules and used as "a priori information" for the next iteration. Several simulation results on typical HDSL-like test-loops confirm the superiority of the proposed turbo-detector (TD) over current solutions based on conventional MMSE decision-feedback equalizers and precoders (such as the Tomlinson-Harashima precoder). The numerical tests also point out that performance of the presented TD on typical high bit-rate digital subscriber lines (HDSL) is not limited by error-floor phenomena, even at error-probabilities below 10/sup -7/.

Motivated by recent high bandwidth communication systems, Inter-Symbol Interference (ISI) channels with 1-bit quantized output are considered under an average-power-constrained continuous input. While the exact capacity is difficult to characterize, an approximation that matches with the exact channel output up to a probability of error is provided. The approximation does not have additive noise, but constrains the channel output (without noise) to be above a threshold in absolute value. The capacity under the approximation is computed using methods involving standard Gibbs distributions. Markovian achievable schemes approaching the approximate capacity are provided. The methods used over the approximate ISI channel result in ideas for practical coding schemes for ISI channels with 1-bit output quantization.

This paper describes a low-latency LDPC- (low-density parity-check) based iterative equalizer, using pre-FFT (time-domain) signal processing, for OFDM (orthogonal frequency division multiplexing) inter-symbol interference (ISI) channels. Combined with a novel channel estimator, the proposed receiver can estimate both channel-state information (CSI) and ISI accurately even under a fast-fading environment. At a data rate of 100 Mbps, and vehicle speed of 100 km/h (at a carrier frequency of f/sub c/=5 GHz), where almost 40% of the OFDM symbols face ISI, the receiver provides a block-error-rate (BLER) of 10/sup 2/ at 0.8-dB degradation from that of an ISI-free environment with perfect CSL. By introducing an adaptive iteration control (AIC) technique, our receiver maintains good performance with a maximum latency of 8 OFDM symbol durations. Computer simulations showed that the proposed scheme is robust and has potential applications in broadband, high-quality communication systems.

In this paper, we introduce a simulation-based method to compute the information rates of intersymbol interference (ISI) channels with signal-dependent Gaussian noise when the inputs are binary and independent identically distributed. The method extends the idea in Arnold et al. (2001), which focuses on the ISI channels with additive white Gaussian noise (AWGN). With the new method, we can compute the information rates of the Lorentzian channels with media noise that represents a suitable model for practical magnetic recording channels.

This paper studies the application of neural networks to Viterbi detection of FTN signals in an intersymbol interference (ISI) channel. The main contribution of this paper is to propose a receiver structure for detecting FTN signals in unknown static ISI channel. In particular, we propose a novel low-complexity neural network structure for calculating the branch metrics, and we explore its suitability for FTN signalling with channel uncertainty. We compare the proposed network, which we call the Metric Net (MetNet), to a benchmark neural network-based technique for metric calculation, the ViterbiNet, which was originally designed for ISI channels. The simulation results confirm that the MetNet outperforms the ViterbiNet, with two orders of magnitude lower complexity, and is much more resilient to channel uncertainty than the traditional Viterbi detector, which uses Euclidean distance for metric calculations. We further show that the MetNet exhibits robustness to being trained at mismatched SNR values and FTN pulse acceleration factors, meaning that the number of trained models required can be significantly reduced. Additionally, the results show that the proposed MetNet remains a favorable alternative at much higher levels of channel uncertainties. The results also reflect that we can generalize the MetNet to work with different channel models defined by different decaying factors. Finally, we show that we succeed in achieving a bandwidth efficiency gain of 33% due to FTN by using the MetNet in the presence of channel uncertainty.

Recently relay networks have attracted a lot of attention for their potential to increase spatial diversity. However, limited attention has been paid to practical detector design and implementation issues. In this paper, we investigate the design of maximum-likelihood detectors for half-duplex amplify-and-forward relay networks in intersymbol interference (ISI) channels. In particular, we study the case when the relay period is long, which results in an effective channel impulse response that is sparse and periodically time-varying. The proposed design provides a good performance/complexity tradeoff.

Rate and diversity impose a fundamental trade-off in communications. This trade-off was investigated for flat-fading channels in as well as for Inter-symbol Interference (ISI) channels in . A different point of view was explored in where high-rate codes were designed so that they have a high-diversity code embedded within them. These diversity embedded codes were investigated for flat fading channels both from an information-theoretic viewpoint and from a coding theory viewpoint in . In this paper, we explore the use of diversity embedded codes for inter-symbol interference channels. In particular the main result of this paper is that the diversity multiplexing trade-off for fading MISO/SIMO/SISO ISI channels is indeed successively refinable. This implies that for fading ISI channels with a single degree of freedom one can embed a high diversity code within a high rate code without any performance loss (asymptotically). This is related to a deterministic structural observation about the asymptotic behavior of frequency response of channel with respect to fading strength of time domain taps as well as a coding scheme to take advantage of this observation.

This paper computes the optimal diversity multiplexing curve for intersymbol interference (ISI) channels and reveal it equals the matched filter bound. A simple transmission scheme that achieves this bound is presented and analyzed. The scheme provides a tractable method of extracting maximal diversity gain for any spatial multiplexing gain, when communicating over multipath fading channels.