In this paper we propose a technique to deal with unknown random jitter in a band-limited Gaussian channel. The jitter caused by the deviation or displacement of signal pulses affects the performance of the communication system. We show that a sampling set as small as twice the baud rate is enough for good detection performance. Detection is done by means of a suboptimal algorithm with polynomial time computational complexity. The results are compared for different transmission rates. The simulation results for Gaussian pulses and sinc pulses show that despite the sub-optimality of our algorithm, the bit-error probability remains relatively close to optimum.

This paper proposes a noise-predictive detection algorithm for intersymbol interference channels with linear regressive noise. A block factorization of the covariance matrix of the linear regressive Gaussian noise is used to derive the branch metrics. This algorithm exhibits near optimal performance. A generalization of this algorithm to signal dependent linear regressive noise is also presented and its performance improvement over conventional algorithms with comparable complexity is shown using simulation results.

We propose a symbol-based belief propagation (BP) algorithm for iterative soft-decision decoding of Reed-Solomon (RS) codes. Complexity reduction is achieved by using a fast Fourier transform (FFT)-based BP algorithm. Parity-check matrix adaptation based on the reliability of the codeword symbols is an essential step to make the BP algorithm effective on high-density parity-check matrices characteristic of RS codes. The matrix adaptation, as well as all other operations, is performed at symbol level such that bit-to-symbol and symbol-to- bit conversions are avoided. A moderate coding gain over algebraic hard-decision decoding is achieved on additive white Gaussian noise channels. The symbol-based algorithm presented in this paper addresses the main weakness of its bit-based counterpart, namely the prohibitively high complexity for practical applications that use long codes and large finite fields, albeit with less coding gain.

We derive a recursive algorithm to compute the joint maximum a posteriori probabilities of p consecutive bits on partial response channels. This channel detector is optimal for nonbinary soft-iterative decoding with multiple channel iterations. It is validated on a nonbinary low-density parity-check coded perpendicular magnetic recording channel.

We propose a soft-decision Reed-Solomon (RS) decoding algorithm, which uses the combined approaches of the Chase and the generalized minimum distance (GMD) algorithms and exploits the recursive property of interpolation-based RS decoding algorithms. We evaluate the performance of this Chase- GMD type algorithm on perpendicular magnetic recording systems, and present simulation results that show that we can achieve better performance with fairly short latency and lower decoding complexity than the current state-of-art algorithms.

Computing maximum-likelihood bounds on the performance of systems involving partial response (PR) channels, with or without an error correcting code present, is rather complex, requiring exponentiating large matrices of polynomials. Being able to practically and efficiently compute these bounds, especially for cases with data sequences and channels more complex than previously considered cases, is expensive in both computational time and memory used. In this paper, we discuss our implementation of a program to compute these bounds for the case of Reed-Solomon codes over PR channels and explore some of the choices and tradeoffs made in that implementation.

We propose efficient forward recursive algorithms for algebraic soft-decision list decoding of Reed-Solomon codes, which utilize channel reliability information, and outperform the Koetter-Vardy (KV) algorithm with lower decoding latency. We evaluate the performance of the proposed decoding algorithms on additive white Gaussian noise and partial response channels. Simulation results show that we can achieve better performance on a modified extended-extended partial response class 4 channel than on the best possible performance of the KV algorithm, as given by the asymptotic bound for high-rate codes.

Type-II hybrid ARQ protocols are widely used in current packet-oriented systems, such as wireless ATM, in combination with powerful error-correction codes for robust transmission and high channel throughput. In this paper, we investigate Reed-Solomon codes and soft-decision decoding algorithms for type-II hybrid ARQ systems, which lead to improved performance and higher channel throughput.

A reduced complexity implementation of a soft Chase algorithm for algebraic soft-decision decoding of Reed-Solomon (RS) codes, based on the recently proposed algorithm of Koetter and Vardy, is presented. The reduction in complexity is obtained at the algorithm level by integrating the re-encoding and Chase algorithms and at the architecture level by considering a backup mode which sharply reduces the average computational complexity of the hybrid decoder.

A method to evaluate the performance of a low-density parity-check (LDPC) code on partial-response (PR) channels in terms of the noise threshold and decoding error is presented. Given a particular codeword or assuming an independent and uniformly distributed (i.u.d.) codeword for transmission, the density-evolution algorithm is used to compute the probability density function of messages passing in the decoding process, from which the decoding error is extracted. This estimated i.u.d. decoding error is used to approximate the decoding error of an ensemble of LDPC codes on arbitrary PR channels. Comparison with simulation results shows that it is a very good approximation for the simulated codes, provided their length is large enough.

Pseudowhitening of oversampled signals in range is proposed as a method to improve the performance of spectral moment and polarimetric variable estimators on weather surveillance radars. In an attempt to overcome the noise sensitivity of the whitening transformation, a solution based on the minimum mean-square-error criterion is considered first; however, this transformation is less practical than whitening because it requires knowledge of the signal-to-noise ratio at every range location. Pseudowhitening techniques are introduced as practical solutions that achieve a suboptimal compromise between variance reduction and noise sensitivity. Based on regularization methods for the solution of ill-conditioned problems, two pseudowhitening schemes are proposed: the clipped singular value decomposition transformation and the sharpening filter. By comparing their statistical performance with theoretical minimum bounds, it is shown that pseudowhitening-based estimators are almost optimal under practical conditions. Estimators based on pseudowhitening techniques avoid the pitfalls of their whitening-transformation-based counterparts and lead to more accurate radar products and/or rapid data acquisition for a much wider range of signal-to-noise ratios.

Signal-to-noise ratio (SNR) mismatch has been found to have a great impact on the performance of low-density parity-check (LDPC) coded magnetic recording channels. While an inappropriate SNR mismatch degrades the performance dramatically, a properly selected optimum SNR mismatch can improve it significantly. We analyze the causes of this phenomenon and find optimum values for the SNR mismatch for specific magnetic recording systems using both density evolution analysis and Monte Carlo simulations.

A soft-decision Reed-Solomon decoding algorithm has been proposed by Koetter and Vardy, which provides a significant coding gain by utilizing channel output reliability information. In this paper we present a Chase-type soft-decision algorithm which provides additional gains at the expense of a small increase in complexity. We evaluate the performance of this decoding algorithm on additive white Gaussian noise channel and Rayleigh fading channels. Simulation results show that coding gains on the order of several dB can be achieved on uncorrelated Rayleigh fading channels over traditional hard-decision Reed-Solomon decoding algorithms.

A soft-decision Reed-Solomon decoding algorithm has been recently proposed by Koetter and Vardy, which provides significant coding gains by fully utilizing channel output reliability information. The algorithm converts this reliability information into a set of interpolation points and their multiplicities, and consists of three day steps: multiplicity matrix calculation, soft interpolation, and polynomial factorization. Large coding gains have been observed only for low-rate codes in additive white Gaussian noise channels, and with very high decoding complexity. Here we present a new suboptimal algorithm for computing the multiplicity matrix, discuss architectures for soft-decision decoders and evaluate the performance of the soft-decision algorithm on magnetic recording channels in the presence of erasures.

This paper discusses a simple yet accurate approach for modeling W-CDMA power amplifiers. Several stages of the modeling process are presented. First, the test signals are defined. W-CDMA signals are approximated using simple periodic signals. These signals are then used as inputs to the power amplifier (PA) measurement system. Second, a process for deriving the AM/AM and AM/PM characteristics is given. The AM/AM and AM/PM characteristics, which are derived directly from the measured PA data, are used as the basis for modeling. The tapped-delay-line (TDL) model is used to model the AM/AM and AM/PM characteristics. The TDL model is accurate to -40 dB when linear filters are inserted before and after the AM/AM and AM/PM characteristics are modeled.

We consider the use of high-rate low-density parity-check (LDPC) codes for magnetic recording. We design and evaluate the performance of a magnetic recording system, which uses an LDPC code as the error-correcting code, in conjunction with a rate 16/17 quasimaximum-transition-run (QMTR) channel code on a modified E/sup 2/PR4 (ME/sup 2/PR4)-equalized channel. Iterative decoding between the partial response channel and the LDPC code is performed. Simulations show that an additional four-dB gain over the QMTR code can be obtained by the LDPC code. The algorithms used to design this LDPC code are also discussed.

The use of base station adaptive antenna arrays is an attractive way to increase the capacity of code division multiple access cellular systems. In this paper, a system with an adaptive minimum redundancy array (MRA) at the base stations is proposed. This system uses the high-resolution signal angle of the arrival estimation algorithm, ESPRIT, in conjunction with array augmentation techniques, and linear least squares adaptation. For the MRA, we propose to use virtual array elements at the locations where there is no sensor element. All real and virtual sensor outputs are weighted and combined to extract the desired signal components and suppress interference. To simplify the evaluation of the interference from adjacent cells, the concentric circle cell geometry is employed in place of the common hexagonal cell geometry. The performance of the proposed system is quantified by comparison with the omnidirectional antenna, and the adaptive uniform linear array (ULA) with the same number of elements and same array aperture, using realistic simulations. It is shown that, for a four-element array, there is about a 4- and a 1-dB improvement in the despread output signal-to-interference-plus-noise ratio of the proposed system over the omnidirectional antenna and the conventional ULA, respectively.

Block turbo codes have been shown to have near-optimum performance and lower decoding complexity than convolutional turbo codes for additive white Gaussian noise (AWGN) channels. In this paper we discuss the applicability of block turbo codes to partial-response (PR) equalized Lorentzian channels for magnetic recording. In particular, we consider the iterative decoding of product codes and parallel concatenated block codes. Simulation results show that both systems offer substantial gains over uncoded systems.

A coding system for magnetic recording channels combining a high rate error detection code with a list Viterbi algorithm (LVA) is proposed. This coding system can provide a coding gain of up to 2-3 dB. An accurate estimation technique of LVA performance is developed.

Steep seismic events are often aliased due to spatial undersampling of wave fields creating different kinds of coherent noises and corrupting the migration process, O. Yilmaz (1987). To avoid the expensive costs of high-resolution spatial sampling, seismic traces are interpolated. A novel spatial interpolation algorithm, based on a narrowband multichannel model, is considered.

Proposes a new recursive version of an earlier technique for fast initialization of data-driven echo cancelers (DDECs). The speed of convergence and the covariance of the estimate of the proposed technique are comparable to the recursive least squares (RLS) algorithm, however, the computational complexity is no greater than the least mean square (LMS) algorithm. Analysis of computational complexity and the estimation error is also provided. Simulation results based on both floating-point and fixed-point arithmetic illustrate a remarkable improvement in terms of speed of convergence and steady-state error over the computationally comparable LMS algorithm.

The use of an adaptive antenna array at the base station in the reverse link of cellular communication systems is proposed. Two kinds of adaptive methods are considered. First, the recursive least squares (RLS) algorithm, which is a reference signal based method, is used to adjust the weight coefficients of the array. For eigenstructure based methods, the ESPRIT algorithm is used for angle of arrival (AOA) estimation and the linear least squares estimate (LLSE) is used to determine the adaptive array weights. The improvement in performance is quantified by comparison with the omnidirectional antenna system using realistic simulations. The comparisons of different adaptive arrays are discussed and analyzed.

The authors propose a new recursive version of a technique proposed by G. Long and F. Ling (1990) for the initialization of a data-driven echo canceler (DDEC). They prove that the Long and Ling algorithm yields a least-squares solution, and then a new technique is presented which is comparable to the recursive-least-squares (RLS) algorithm. However, the use of a unique training sequence reduces the complexity of the RLS algorithm to that of the least-mean-square (LMS) algorithm. An analysis of the covariance of the estimated weight vector is presented, and simulation results show a remarkable improvement in both convergence speed and steady-state error.>

Statistical analysis of a fast algorithm for frequency estimation is presented. Details of how the analysis was derived with the aid of a symbolic algebra software package, Maple, are given, and some issues involved in the use of symbolic algebra packages in signal processing research are discussed.>

The measurement of clear-air turbulence with a Doppler radar is investigated. An autoregressive moving average (ARMA) model is proposed to improve the Doppler spectral width estimates. An iterative algorithm that has its origin in system identification is used for the estimation of the ARMA parameters. By taking advantage of a priori knowledge of the correlation matrix, which arises in the derivation of the governing equations of the ARMA parameters, the ARMA spectral estimate can be improved. This improvement is shown in terms of bias and variance of the spectral width estimate.>

A unified approach to three eigendecomposition-based methods for frequency estimation in the presence of noise is presented. These are the Tufts-Kumaresan (TK) method, the minimum-norm (MN) method, and the total-least-squares (TLS) method. It is shown that the MN method is a modified version of the TK method and the TLS method is a generalization of the MN methods. The TLS solution vector is expressed in matrix form, and an alternate way of computing it is presented. The MN methods exhibit some improvement over the TK method.>

A new class of fast least squares linear prediction algorithms for frequency estimation is described, and a specific example of this class of algorithms is presented in detail. A simulation for the case of single-sinusoid inputs shows that the estimates are quite sensitive to noise for frequencies near half the Nyquist frequency. A simplified error analysis seems to confirm the existence of the noise sensitivity in that range of frequencies.>

The performance of minimum-shift keying (MSK) and sinusoidal frequency-shift keying (SFSK) in adjacent-channel interference and frequency-selective fading is investigated analytically. The receiver system considered uses a sampling detector and hence is suboptimum. In addition to the above impairments, the system model incorporates other disturbances, namely stationary white Gaussian noise, quadrature channel interference, random timing misalignment, and random carrier phases of interfering signals. Results and comparisons are presented for both coupled effects and individual effects of frequency-selective fading, adjacent-channel interference, and quadrature-channel interference.>

Linear-prediction-based methods are applied to mean Doppler estimation of radar signals in the presence of ground clutter when only a small number of samples is available for processing. The authors show the ability of these methods to resolve weather signals from ground clutter in unfavourable conditions, i.e. when the clutter is stronger and its spectrum is narrower than the signal's. It is shown that the Prony method can be applied to this task. In addition, given the a priori information about ground clutter, the Prony method can be simplified to derive a convenient formula for mean Doppler frequency estimation in the presence of ground clutter. This estimator can be considered as a generalization of the pulse-pair (PP) estimator. Such a scheme can give satisfactory results in resolving weather signals from ground clutter for signal-to-noise ratios (SNR) above 20 dB. The eigendecomposition-based minimum-norm method is applied to lower the SNR threshold to 10 dB. The methods are compared with combinations of various ground clutter filters and the PP estimator.>

A statistical analysis of the estimates of the damping factor of a single complex exponential in additive white Gaussian noise is presented. The analysis is done for two of the more popular methods, namely, the system identification method of principal Hankel components (PHC) and the Tufts-Kumaresan (TK) method of linear prediction. Assuming a high signal-to-noise ratio, closed form expressions for the variances of the damping factor are derived. These analytical solutions are confirmed with computer simulations. The analysis indicates that both the TK method and the PHC method perform well in comparison to the Cramer-Rao bound. Theoretically, however, the PHC method slightly outperforms the TK method.>

A complex form of a fast frequency estimation technique based on the quotient-difference (QD) algorithm is presented. The complex stochastic process is modeled as a sum of nonharmonically related damped exponentials. This algorithm computes the zeros of the predictor polynomial in a single step, directly from the data and has computational complexity on the order of Np.>

This paper presents a fast method to compute high resolution frequency estimates. The technique is based on the theory of Rutishouser polynomials. It is shown that the quotient-difference algorithm can be used to accurately compute the frequencies directly from the data. An example is given that illustrates the algorithm's performance on a standard data set.

A spectral estimation technique based on the quotient-difference (QD) algorithm of Rutishouser is proposed. The model assumed is a sum of nonharmonically related damped exponentials. The frequencies and damping factors are found directly from the data avoiding the need for rooting a polynomial as required in the closely related Prony' s Energy Spectral Density (ESD) technique.

The performances of two low-complexity IIR adaptive filtering algorithms applied to the equalization of a typical transmission line channel are compared. The algorithms considered are the Simple Hyperstable Adaptive Recursive Filter (SHARF) and the stochastic gradient equation error algorithm. Gradient descent techniques are shown to be ill suited for transmission line channels.