Search

We present a general nonlinear Bayesian filter for high-dimensional state estimation using the theory of reproducing kernel Hilbert space (RKHS). By applying the kernel method and the representer theorem to perform linear quadratic estimation in a functional space, we derive a Bayesian recursive state estimator for a general nonlinear dynamical system in the original input space. Unlike existing nonlinear extensions of the Kalman filter where the system dynamics are assumed known, the state-space representation for the Functional Bayesian Filter (FBF) is completely learned online from measurement data in the form of an infinite impulse response (IIR) filter or recurrent network in the RKHS, with universal approximation property. Using a positive definite kernel function satisfying Mercer’s conditions to compute and evolve information quantities, the FBF exploits both the statistical and time-domain information about the signal, extracts higher-order moments, and preserves the properties of covariances without the ill effects due to conventional arithmetic operations. We apply this novel kernel adaptive filtering (KAF) to recurrent network training, chaotic time-series estimation and cooperative filtering using Gaussian and non-Gaussian noises, and inverse kinematics modeling. Simulation results show FBF outperforms existing Kalman-based algorithms. [ABSTRACT FROM AUTHOR]

Principal component analysis (PCA) is a classical and ubiquitous method for reducing data dimensionality, but it is suboptimal for heterogeneous data that are increasingly common in modern applications. PCA treats all samples uniformly so degrades when the noise is heteroscedastic across samples, as occurs, e.g., when samples come from sources of heterogeneous quality. This paper develops a probabilistic PCA variant that estimates and accounts for this heterogeneity by incorporating it in the statistical model. Unlike in the homoscedastic setting, the resulting nonconvex optimization problem is not seemingly solved by singular value decomposition. This paper develops a heteroscedastic probabilistic PCA technique (HePPCAT) that uses efficient alternating maximization algorithms to jointly estimate both the underlying factors and the unknown noise variances. Simulation experiments illustrate the comparative speed of the algorithms, the benefit of accounting for heteroscedasticity, and the seemingly favorable optimization landscape of this problem. Real data experiments on environmental air quality data show that HePPCAT can give a better PCA estimate than techniques that do not account for heteroscedasticity. [ABSTRACT FROM AUTHOR]

One of the most important problems in the estimation theory is the one-step prediction. The goal of this problem is to determine the predictions of states in the next time step. This paper focuses on the one-step prediction for nonlinear dynamic systems. The system under investigation involves unknown inputs and the system noises are correlated. In this approach, using the Taylor series expansion for nonlinear functions, a new augmented state nonlinear predictor is proposed for discrete time-varying nonlinear systems. This predictor is obtained by solving a deterministic min-max optimization based on the regularized least squares problem. Moreover, to reduce the computational complexity of the prediction solution, using a nonlinear transformation, we propose a two-stage predictor including lower order estimators for states and unknown inputs. Finally, a frequency modulated signal model is considered to illustrate the effectiveness and performance of the proposed approaches in comparison with the existing estimation methods. [ABSTRACT FROM AUTHOR]

Upper and lower bounds on the minimum mean square error for additive noise channels are derived when the input distribution is constrained to be close to a Gaussian reference distribution in terms of the Kullback–Leibler divergence. The upper bound is tight and is attained by a Gaussian distribution whose mean is identical to that of the reference distribution and whose covariance matrix is defined implicitly via a system of non-linear equations. The estimator that attains the upper bound is identified as a minimax optimal estimator that is robust against deviations from the assumed prior. The lower bound provides an alternative to well-known inequalities in estimation and information theory—such as the Cramér–Rao lower bound, Stam's inequality, or the entropy power inequality—that is potentially tighter and defined for a larger class of input distributions. Several examples of applications in signal processing and information theory illustrate the usefulness of the proposed bounds in practice. [ABSTRACT FROM AUTHOR]

We consider the problem of forecasting a high-dimensional time series that can be modeled as matrices where each column denotes a measurement and use low-rank matrix factorization for predicting future values or imputing missing ones. We define and analyze our problem in the online setting in which the data arrive as a stream and only a single pass is allowed. We present and analyze new matrix factorization techniques that can learn low-dimensional embeddings effectively in an online manner. Based on these embeddings, we derive a recursive minimum mean square error estimator based on an autoregressive model. Experiments with two real datasets of tens of millions of measurements show the benefits of the proposed approach. [ABSTRACT FROM AUTHOR]

We consider the matrix completion problem with a deterministic pattern of observed entries. In this setting, we aim to answer the question: Under what conditions will there be (at least locally) unique solution to the matrix completion problem, i.e., the underlying true matrix is identifiable? We answer the question from a certain point of view and outline a geometric perspective. We give an algebraically verifiable sufficient condition, which we call the well-posedness condition, for the local uniqueness of minimum rank matrix completion (MRMC) solutions. It appears that this condition is basic for the analysis of MRMC, and we show that, in a sense, the condition is generic. We also argue that the low-rank approximation approaches are more stable than MRMC and further propose a sequential statistical testing procedure to determine the “true” rank from observed entries. Finally, we provide numerical examples aimed at verifying validity of the presented theory. [ABSTRACT FROM AUTHOR]

Detection of the number of source signals (NoSS) in the presence of additive noise is considered. We present a new approach denoted by the mean squared eigenvalue error (MSEE). The MSEE is the mean squared error between the desired noise-free eigenvalues and the available estimated eigenvalues. The approach investigates and analyzes the probabilistic distribution of the available eigenvalue estimates and revisits proper thresholding of these sorted values. The optimum NoSS is provided by minimizing the MSEE. A probabilistic worst-case technique is proposed to estimate the value of the MSEE by using only the available data. It is shown that the proposed method is consistent as the data length increases. It is also shown that the method is consistent as the signal-to-noise ratio (SNR) increases. Simulation results illustrate advantages of the MSEE over competing approaches and confirm effectiveness and robustness of the MSEE even in low-SNR or small sample size scenarios. [ABSTRACT FROM AUTHOR]

Many state-of-the-art signal decomposition techniques rely on a low-rank factorization of a time–frequency (t–f) transform. In particular, nonnegative matrix factorization (NMF) of the spectrogram has been considered in many audio applications. This is an analysis approach in the sense that the factorization is applied to the squared magnitude of the analysis coefficients returned by the t–f transform. In this paper, we instead propose a synthesis approach, where low-rankness is imposed to the synthesis coefficients of the data signal over a given t–f dictionary (such as a Gabor frame). As such, we offer a novel modeling paradigm that bridges t–f synthesis modeling and traditional analysis based NMF approaches. The proposed generative model allows us in turn to design more sophisticated multilayer representations that can efficiently capture diverse forms of structure. Additionally, the generative modeling allows us to exploit t–f low-rankness for compressive sensing. We present efficient iterative shrinkage algorithms to perform estimation in the proposed models and illustrate the capabilities of the new modeling paradigm over audio signal processing examples. [ABSTRACT FROM AUTHOR]

In this paper, the problem of estimating the unknown degree of sparsity from compressive measurements without the need to carry out a sparse recovery step is investigated. While the sparsity order can be directly inferred from the effective rank of the observation matrix in the multiple snapshot case, this appears to be impossible in the more challenging single snapshot case. It is shown that specially designed measurement matrices allow to rearrange the measurement vector into a matrix such that its effective rank coincides with the effective sparsity order. In fact, it is proven that matrices that are composed of a Khatri–Rao product of smaller matrices generate measurements that allow to infer the sparsity order. Moreover, if some samples are used more than once, one of the matrices needs to be Vandermonde. These structural constraints reduce the degrees of freedom in choosing the measurement matrix, which may incur in a degradation in the achievable coherence. Thus, this paper also addresses suitable choices of the measurement matrices. In particular, Khatri–Rao and Vandermonde matrices are analyzed in terms of their coherence and a new design for Vandermonde matrices that achieves a low coherence is proposed. [ABSTRACT FROM AUTHOR]

This paper focuses on the linear optimal recursive sequential fusion filter design for multisensor systems subject to stochastic parameter perturbations, fading measurements, and correlated noises. The stochastic parameter perturbations existing in the state model are described by white multiplicative noises. The fading measurement phenomena for different sensors are described by independent random variables with known statistical properties. Moreover, the measurement noises of different sensors are correlated with each other and also correlated with the system noise at the same time step. First, a model equivalent to the original system is established by transferring the multiplicative noises into the additive noises. Then, based on the equivalent model and an innovation analysis method, a sequential fusion filter in the linear minimum variance sense is proposed to solve the linear optimal state estimation problem in real time according to the arriving order of measurements from different sensors. Finally, the equivalence on estimation accuracy of the proposed sequential fusion filter and the centralized fusion filter is strictly proven, which shows the optimality of the proposed sequential fusion algorithm. Moreover, the proposed sequential fusion filter has a reduced computational burden. Compared with the distributed matrix-weighted fusion filter, the computation of cross-covariance matrices is avoided and the estimation accuracy is improved. Finally, a simulation example verifies the effectiveness of the proposed sequential fusion filtering algorithm. [ABSTRACT FROM AUTHOR]

Time Difference of Arrival (TDOA) positioning has been an active research over the years due to its rich amounts of applications. Improving the quality of the TDOA measurements, especially under harsh scenarios, is crucial to achieve better localization accuracy and higher noise tolerance. Increasing the TDOA quality has been traditionally based on the statistical estimation point of view. Nevertheless, TDOA exhibits interesting algebraic properties, which when exploited can further enhance the measurements. This paper develops two algebraic properties of TDOA in the situation where multiple sources and sensors are present, such as sensor network and joint source-sensor localizations. The algebraic properties are new and have not appeared in the literature. They are in terms of the rank behaviors of two matrices that are formed by the TDOA themselves and their squared values. The development starts from an established property of Time of Arrival and applies determinant and vector space manipulations to obtain the concrete results. An example is included to illustrate the use of the algebraic properties for reducing the noise in TDOAs. [ABSTRACT FROM AUTHOR]

This two-part paper considers an uplink massive device communication scenario in which a large number of devices are connected to a base station (BS), but user traffic is sporadic so that in any given coherence interval, only a subset of users is active. The objective is to quantify the cost of active user detection and channel estimation and to characterize the overall achievable rate of a grant-free two-phase access scheme in which device activity detection and channel estimation are performed jointly using pilot sequences in the first phase and data is transmitted in the second phase. In order to accommodate a large number of simultaneously transmitting devices, this paper studies an asymptotic regime where the BS is equipped with a massive number of antennas. The main contributions of Part I of this paper are as follows. First, we note that as a consequence of having a large pool of potentially active devices but limited coherence time, the pilot sequences cannot all be orthogonal. However, despite the nonorthogonality, this paper shows that in the asymptotic massive multiple-input multiple-output regime, both the missed device detection and the false alarm probabilities for activity detection can always be made to go to zero by utilizing compressed sensing techniques that exploit sparsity in the user activity pattern. Part II of this paper further characterizes the achievable rates using the proposed scheme and quantifies the cost of using nonorthogonal pilot sequences for channel estimation in achievable rates. [ABSTRACT FROM AUTHOR]

Detecting periodicity in a short sequence is an important problem, with many applications across science and engineering. Several efficient algorithms have been proposed for this over the years. There is a wide choice available today in terms of the tradeoff between algorithmic complexity and estimation accuracy. In spite of such a rich history, one particular aspect of period estimation has received very little attention from a fundamental perspective. Namely, given a discrete time periodic signal and a list of candidate integer periods, what is the absolute minimum datalength required to estimate its integer period? Notice that the answer we seek must be a fundamental bound, i.e., independent of any particular period estimation technique. Common intuition suggests the minimum datalength as twice the largest expected period. However, this is true only under some special contexts. This paper derives the exact necessary and sufficient bounds to this problem. The above-mentioned question is also extended to the case of mixtures of periodic signals. First, a careful mathematical formulation discussing the unique identifiability of the component periods (hidden integer periods) is presented. Once again, a rigorous theoretical framework in this regard is missing in the existing literature but is a necessary platform to derive precise bounds on the minimum necessary datalength. The bounds given here are generic, that is, independent of the algorithms used. Specific algorithm-dependent bounds are also presented in the end for the case of dictionary-based integer period estimation reported in recent years. [ABSTRACT FROM AUTHOR]

The use of channel code is mandatory in current digital communication systems. It allows us to access the information on the receiver side despite the presence of noise. In this paper, we are interested in the blind identification of the parameters of an error correcting code from a received noisy data stream. The literature provides a large amount of contributions for this problem in the hard-decision case but few in the soft-decision case. It is well known that soft-decision methods allow significant gain in decoding techniques. Thence, we propose an algorithm which is able to identify the length of a code through a classification process from the bits likelihood values. It highlights a difference of behavior between an independent identically distributed sequence and an encoded one. This method does not rely on any a priori knowledge about the encoder involved. Indeed, the distribution of $n$ -length code words in an $n$ -dimensional space depends on the encoder characteristics. Some areas of this $n$ -dimensional space are left vacant because of the redundancy added by the encoder. Despite the presence of noise, it is still possible to detect this phenomenon. Furthermore, an adaptation of a collisions method based on the birthday paradox gives us access to an estimation of the code dimension. Finally, we investigate the performance of this estimation methods to show their efficiency. [ABSTRACT FROM AUTHOR]

We propose a novel soft-input soft-output equalizer for single-carrier transmissions over unknown frequency-selective block-fading channels. Our equalizer leverages the recently proposed parametric bilinear generalized approximate message passing algorithm for joint channel-estimation and symbol-detection, and exploits fast Fourier transform (FFT)-processing to achieve a per-symbol complexity that grows only logarithmically in the channel delay-spread. Furthermore, it supports the use of Gaussian mixture models to support the approximately sparse nature of wideband wireless channel responses. Numerical experiments, conducted using physically motivated Saleh–Valenzuela channel models, show that the proposed approach achieves channel normalized mean square error and bit error rate that are significant improved over existing turbo frequency-domain equalization approaches for unknown channels. Additional experiments show that the proposed scheme facilitates much higher spectral efficiencies than sparse deconvolution methods based on convex relaxation. [ABSTRACT FROM AUTHOR]

Time-division duplex (TDD)-based massive MIMO systems rely on the reciprocity of the wireless propagation channels when calculating the downlink precoders based on uplink pilots. However, the effective uplink and downlink channels incorporating the analog radio front-ends of the base station (BS) and user equipments (UEs) exhibit non-reciprocity due to non-identical behavior of the individual transmit and receive chains. When the downlink precoder is not aware of such channel non-reciprocity (NRC), system performance can be significantly degraded due to the NRC-induced interference terms. In this paper, we consider a general TDD-based massive MIMO system where frequency-response mismatches at both the BS and UEs, as well as the mutual coupling mismatches at the BS large-antenna system all coexist and induce channel NRC. Based on the NRC-impaired signal models, we first propose a novel iterative estimation method for acquiring both the BS and UE side NRC matrices and then also propose an efficient NRC-aware downlink precoder design which utilizes the obtained estimates. Furthermore, an efficient pilot signaling scheme between the BS and UEs is introduced in order to facilitate executing the proposed estimation method and the NRC-aware precoding technique in practical systems. Comprehensive numerical results indicate substantially improved spectral efficiency performance when the proposed NRC estimation and NRC-aware precoding methods are adopted, compared to the existing state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Wideband direction of arrival (DOA) estimation with sensor arrays is an essential task in sonar, radar, acoustics, biomedical, and multimedia applications. Many state-of-the-art wideband DOA estimators coherently process frequency binned array outputs by approximate maximum likelihood (ML), weighted subspace fitting, or focusing techniques. This paper shows that bin signals obtained by filter-bank approaches do not obey the finite rank narrow-band array model, because spectral leakage and changes of the array response with frequency within the bin create ghost sources dependent on the particular realization of the source process. Therefore, existing DOA estimators based on binning are not consistent even when the array response is perfectly known. In this paper, under a more realistic array model, which still has finite rank under a space-time formulation, signal subspaces at arbitrary frequencies can be consistently recovered under mild conditions by applying space time, MUSIC-type (ST-MUSIC) estimators to the dominant eigenvectors of the wideband, space-time sensor cross-correlation matrix. A novel, consistent ML-based ST-MUSIC subspace estimate is developed to estimate the number of sources active at each frequency by information theoretic criteria. Empirical ST-MUSIC subspaces are fed to any subspace fitting DOA estimator at single or multiple frequencies. Simulations confirm that this approach allows better performance over binning approaches at sufficiently high signal to noise ratio, when model mismatches exceed the noise floor. [ABSTRACT FROM AUTHOR]

This paper addresses the problem of downlink channel estimation in frequency-division duplexing massive multiple-input multiple-output systems. The existing methods usually exploit hidden sparsity under a discrete Fourier transform (DFT) basis to estimate the downlink channel. However, there are at least two shortcomings of these DFT-based methods: first, they are applicable to uniform linear arrays (ULAs) only, since the DFT basis requires a special structure of ULAs; and second, they always suffer from a performance loss due to the leakage of energy over some DFT bins. To deal with the above-mentioned shortcomings, we introduce an off-grid model for downlink channel sparse representation with arbitrary two-dimensional-array antenna geometry, and propose an efficient sparse Bayesian learning approach for the sparse channel recovery and off-grid refinement. The main idea of the proposed off-grid method is to consider the sampled grid points as adjustable parameters. Utilizing an in-exact block majorization–minimization algorithm, the grid points are refined iteratively to minimize the off-grid gap. Finally, we further extend the solution to uplink-aided channel estimation by exploiting the angular reciprocity between downlink and uplink channels, which brings enhanced recovery performance. [ABSTRACT FROM AUTHOR]

A number of recent works have proposed to solve the line spectral estimation problem by applying off-the-grid extensions of sparse estimation techniques. These methods are preferable over classical line spectral estimation algorithms because they inherently estimate the model order. However, they all have computation times that grow at least cubically in the problem size, thus limiting their practical applicability in cases with large dimensions. To alleviate this issue, we propose a low-complexity method for line spectral estimation, which also draws on ideas from sparse estimation. Our method is based on a Bayesian view of the problem. The signal covariance matrix is shown to have Toeplitz structure, allowing superfast Toeplitz inversion to be used. We demonstrate that our method achieves estimation accuracy at least as good as current methods and that it does so while being orders of magnitudes faster. [ABSTRACT FROM AUTHOR]

Multicarrier wireless communication systems, which usually operate over frequency-selective fading channels, are practically also impaired by environmental impulsive noise. In order to boost the signal-to-noise power ratio (SNR) at the receiver, this paper proposes and analyzes nonlinear estimators based on the multiple thresholding, with associated piecewise attenuation, or clipping, of the received signal amplitude. The proposed approach exploits a new heuristic criterion to set the thresholds. Although the obtained thresholds are slightly suboptimal in output SNR, the proposed approach has the nice and useful property to allow closed-form analytical derivations for both the threshold(s) and the associated attenuating/clipping parameters, when a Gaussian source is impaired by an impulsive noise. Such a quite general class of estimators, which could be applied also in other scenarios, is particularly attractive for orthogonal frequency division multiplexing communication systems in the presence of additive multicomponent impulsive noise. We are also letting to derive closed-form expressions for the output SNR in frequency-selective multipath fading channels. The SNR performance and the associated symbol error rate have been compared with those of more traditional blanking, clipping, and clipping-blanking processors. Specifically, analytical and simulation results, carried out for Class-A and symmetrical alpha-stable (SαS) impulsive noise, show that the proposed threshold-based suppressors are superior to the traditional ones. Furthermore, as the number of thresholds increases, the proposed estimators closely approach the performance of the optimal minimum mean square error Bayesian estimator. However, as it is shown, in practical conditions only few thresholds are necessary. [ABSTRACT FROM AUTHOR]

This paper considers the detection of a distributed target in a partial observation scenario. A distributed targetmodel is usually adopted when the target size is larger than the range bin, for example, in some high-resolution radars. Based on the distributed target model, joint processing of several consecutive range bins can be adopted to achieve performance improvement with respect to processing just one range bin. In applications in complex electromagnetic environments, the radars may miss some of their observations, a phenomenon that usually caused by interference, spectrum sharing, and so on. This partial observation problem leads to degradation of the estimation accuracy for the disturbance (clutter plus noise) covariance matrix and target amplitude vector, which results in decline of the target detection performance. In this paper, a scheme is proposed by using the low-rank priori knowledge of clutter covariance matrix to estimate the detector's unknown parameters. Specifically, the target amplitude vector is obtained by maximizing the likelihood function, and the disturbance covariance matrix is reconstructed by solving an optimization that considers both the likelihood maximization and low-rank property of the clutter covariance matrix. The simulation results indicate that this algorithm improves the estimation accuracy, and achieves a better detection performance in the partial observation scenarios. [ABSTRACT FROM AUTHOR]

We propose a data-driven algorithm for the Bayesian estimation of stochastic processes fromnoisy observations. The primary statistical properties of the sought signal are specified by the penalty function (i.e., negative logarithm of the prior probability density function). Our alternating direction method of multipliers (ADMM) based approach translates the estimation task into successive applications of the proximal mapping of the penalty function. Capitalizing on this direct link, we define the proximal operator as a parametric spline curve and optimize the spline coefficients by minimizing the average reconstruction error for a given training set. The key aspects of our learning method are that the associated penalty function is constrained to be convex and the convergence of the ADMM iterations is proven. As a result of these theoretical guarantees, adaptation of the proposed framework to different levels of measurement noise is extremely simple and does not require any retraining. We apply our method to estimation of both sparse and nonsparse models of Lévy processes for which the minimum mean square error (MMSE) estimators are available. We carry out a single training session for a fixed level of noise and perform comparisons at various signal-to-noise ratio values. Simulations illustrate that the performance of our algorithm are practically identical to the one of the MMSE estimator irrespective of the noise power. [ABSTRACT FROM AUTHOR]

In many applications, sensors that map the positions of objects in unknown environments are installed on dynamic platforms. As measurements are relative to the observer's sensors, scene mapping requires accurate knowledge of the observer state. However, in practice, observer reports are subject to positioning errors. Simultaneous localization and mapping addresses the joint estimation problem of observer localization and scene mapping. State-of-the-art approaches typically use visual or optical sensors and therefore rely on static beacons in the environment to anchor the observer estimate. However, many applications involving sensors that are not conventionally used for Simultaneous Localization and Mapping (SLAM) are affected by highly dynamic scenes, such that the static world assumption is invalid. This paper proposes a novel approach for dynamic scenes, called GEneralized Motion (GEM) SLAM. Based on probability hypothesis density filters, the proposed approach probabilistically anchors the observer state by fusing observer information inferred from the scene with reports of the observer motion. This paper derives the general, theoretical framework for GEM-SLAM, and shows that it generalizes existing Probability Hypothesis Density (PHD)-based SLAM algorithms. Simulations for a model-specific realization using range-bearing sensors and multiple moving objects highlight that GEM-SLAM achieves significant improvements over three benchmark algorithms. [ABSTRACT FROM AUTHOR]

In this paper, we derive a novel batch Bayesian weighted instrumental variable estimator for the three-dimensional (3D) target motion analysis (TMA) problem using bearing and elevation measurements. Unlike most existing 3D estimators based on instrumental variables, the proposed approach is able to incorporate a priori information into the estimation process and is proven to be approximately asymptotically unbiased. An approximate asymptotic covariance matrix is also presented to evaluate the performance of the estimator. Simulation results show that the proposed estimator outperforms its non-Bayesian counterpart and has an estimation performance on par with the conventional iterative maximum a posteriori estimator, while being orders of magnitude faster in a passive sonar TMA problem. Moreover, it was shown that the proposed approach can provide an accurate initialization for recursive estimators. [ABSTRACT FROM AUTHOR]

A new class of disturbance covariance matrix estimators for radar signal processing applications is introduced following a geometric paradigm. Each estimator is associated with a given unitary invariant norm and performs the sample covariance matrix projection into a specific set of structured covariance matrices. Regardless of the considered norm, an efficient solution technique to handle the resulting constrained optimization problem is developed. Specifically, it is shown that the new family of distributionfree estimators shares a shrinkage-type form; besides, the eigenvalues estimate just requires the solution of a one-dimensional convex problem whose objective function depends on the considered unitary norm. For the two most common norm instances, i.e., Frobenius and spectral, very efficient algorithms are developed to solve the aforementioned one-dimensional optimization leading to almost closed-form covariance estimates. At the analysis stage, the performance of the new estimators is assessed in terms of achievable signal-to-interference-plus-noise ratio (SINR) both for spatial and Doppler processing scenarios assuming different data statistical characterizations. The results show that interesting SINR improvementswith respect to some counterparts available in the open literature can be achieved especially in training starved regimes. [ABSTRACT FROM AUTHOR]

We propose a novel nearest-neighbors approach to organize and curb the growth of radial basis function network in kernel adaptive filtering (KAF). The nearest-instance-centroid-estimation (NICE) kernel least-mean-square (KLMS) algorithm provides an appropriate time-space tradeoff with good performance. Its centers in the input/feature space are organized by quasi-orthogonal regions for greatly simplified filter evaluation. Instead of using all centers to evaluate/update the function approximation at every new point, a linear search among the iteratively-updated centroids determines the partial function to be used, naturally forming locally-supported partial functionals. Under this framework, partial functionals that compose the adaptive filter are quickly stored/retrieved based on input, each corresponding to a specialized “spatial-band” subfilter. The filter evaluation becomes the update of one of the subfilters, creating a content addressable filter bank (CAFB). This CAFB is incrementally updated for new signal applications with mild constraints, always using the past-learned partial filter sums, opening the door for transfer learning and significant efficiency for new data scenarios, avoiding training from scratch as have been done since the invention of adaptive filtering. Using energy conservation relation, we show the sufficient condition for mean square convergence of the NICE-KLMS algorithm and establish the upper and lower bounds of steady-state excess-mean-square-error (EMSE). Simulations on chaotic time-series prediction demonstrate similar levels of accuracy as existing methods, but with much faster computation involving fewer input samples. Simulations on transfer learning using both synthetic and real-world data demonstrate that NICE CAFB can leverage previously learned knowledge to related task or domain. [ABSTRACT FROM AUTHOR]

Polar codes have gained significant amount of attention during the past few years and have been selected as a coding scheme for the next generation of mobile broadband standard. Among decoding schemes, successive-cancellation list (SCL) decoding provides a reasonable tradeoff between the error-correction performance and hardware implementation complexity when used to decode polar codes, at the cost of limited throughput. The simplified SCL (SSCL) and its extension SSCL-SPC increase the speed of decoding by removing redundant calculations when encountering particular information and frozen bit patterns (rate one and single parity check codes), while keeping the error-correction performance unaltered. In this paper, we improve SSCL and SSCL-SPC by proving that the list size imposes a specific number of path splitting required to decode rate one and single parity check codes. Thus, the number of splitting can be limited while guaranteeing exactly the same error-correction performance as if the paths were forked at each bit estimation. We call the new decoding algorithms Fast-SSCL and Fast-SSCL-SPC. Moreover, we show that the number of path forks in a practical application can be tuned to achieve desirable speed, while keeping the error-correction performance almost unchanged. Hardware architectures implementing both algorithms are then described and implemented: It is shown that our design can achieve \mathbf 1.86 Gb/s throughput, higher than the best state-of-the-art decoders. [ABSTRACT FROM PUBLISHER]

In this paper, the problem of estimating complex-valued proportional covariance matrices is addressed. The obtained estimate generalizes the fixed point estimate to scaled packets of data, and is hence called the generalized fixed point estimate (GFPE). The statistical properties (bias, consistency, and asymptotical distributions) of the estimate are presented, and verified through simulations. As an example of application, a radar detection problem is considered. The GFPE is used in the well-known adaptive normalized matched filter (ANMF), and the obtained empirical probability of false alarms threshold curve is compared to the theoretical one. The ANMF is also used to compare the GFPE to the fixed point estimate. [ABSTRACT FROM PUBLISHER]

Adaptive sampling theory has shown that, with proper assumptions on the signal class, algorithms exist to reconstruct a signal in \mathbb R^d with an optimal number of samples. We generalize this problem to the case of spatial signals, where the sampling cost is a function of both the number of samples taken and the distance traveled during estimation. This is motivated by our work studying regions of low oxygen concentration in the Great Lakes. We show that for one-dimensional threshold classifiers, a tradeoff between the number of samples taken and distance traveled can be achieved using a generalization of binary search, which we refer to as quantile search. We characterize both the estimation error after a fixed number of samples and the distance traveled in the noiseless case, as well as the estimation error in the case of noisy measurements. We illustrate our results in both simulations and experiments and show that our method outperforms existing algorithms in a large range of sampling scenarios. [ABSTRACT FROM AUTHOR]

Time-warped (TW) almost-cyclostationary (ACS) processes are obtained from ACS processes by linear time-variant transformations that modify the time scale. TW-ACS processes are shown to be a subclass of the oscillatory ACS processes, a recently introduced class of signals extending that of the ACS ones. The TW-ACS model is useful to describe signals with hidden imperfect periodicities. An example is the received signal when the transmitted one is ACS in the case of general motion law between transmitter and receiver. Some biological signals are other examples. The problem of statistical function estimation is addressed and two estimation techniques are proposed. The theoretical results are corroborated by numerical experiments on simulated and real data. [ABSTRACT FROM PUBLISHER]

We consider the problem of estimating an unknown parameter that is observed indirectly by a sensor network. Each of the sensors exhibits probabilistic nonlinear behavior causing censoring (clipping) of the signals at random. The signals then get transmitted over fading wireless channels to the fusion center. We develop three algorithms for estimating the unknown parameter: first is based on Pearson's Method of Moments, which involves the calculation of the statistical moments of the model that can be derive exactly; second is based on the Maximum-Likelihood Estimator, which results in an intractable likelihood function. To evaluate the likelihood function, we develop a novel approximation via a nonparametric probability density estimator that is based on series expansions of the Gram–Charlier family of basis functions; and third is the Marginalized Least Squares in which we solve the least squares problem by marginalizing out the random unknown nuisance parameters of the model. All three algorithms enjoy a low computational complexity, and only involve solving a one-dimensional optimization problem that is simple to implement in practice. We compare the performance of the proposed algorithms under various system configurations and show various tradeoffs between the algorithms as a function of system parameters (e.g., number of sensors, frame length, signal-to-noise ratio, sensing quality etc.). [ABSTRACT FROM PUBLISHER]

Diffusion adaptive networks tasked with solving estimation problems have attracted attention in recent years due to their reliability, scalability, resource efficiency, and resilience to node and link failure. Diffusion adaptation strategies that are based on the least-mean-squares algorithm can be nonrobust against impulsive noise corrupting the measurements. Impulsive noise can degrade stability and steady-state performance, leading to unreliable estimates. In previous work [“Robust adaptation in impulsive noise,” IEEE Trans. Signal Process., vol. 64, no. 11, pp. 2851–2865, Jun. 2016], a robust adaptive algorithm for stand-alone agents was developed, one that semi-parametrically estimates the optimal error nonlinearity jointly with the parameter of interest. Prior knowledge of the impulsive noise distribution was not assumed. In this paper, we extend the framework to solve the problem of robust distributed estimation by a network of agents. Challenges arise due to the coupling among the agents and the distributed nature of the problem. The resulting diffusion strategy is analyzed and its performance illustrated by numerical simulations. [ABSTRACT FROM PUBLISHER]

Time-varying mixture models are useful for representing complex, dynamic distributions. Components in the mixture model can appear and disappear, and persisting components can evolve. This allows great flexibility in streaming data applications where the model can be adjusted as new data arrives. Fitting a mixture model with computational guarantees which can meet real-time requirements is challenging with existing algorithms, especially when the model order can vary with time. Existing approximate inference methods may require multiple restarts to search for a good local solution. Monte-Carlo methods can be used to jointly estimate the model order and model parameters, but when the distribution of each mixand has a high-dimensional parameter space, they suffer from the curse of dimensionality and and from slow convergence. This paper proposes a generative model for time-varying mixture models, tailored for mixtures of discrete-time Markov chains. A novel, deterministic inference procedure is introduced and is shown to be suitable for applications requiring real-time estimation, and the method is guaranteed to converge at each time step. As a motivating application, we model and predict traffic patterns in a transportation network. Experiments illustrate the performance of the scheme and offer insights regarding tuning of the algorithm parameters. The experiments also investigate the predictive power of the proposed model compared to less complex models and demonstrate the superiority of the mixture model approach for prediction of traffic routes in real data. [ABSTRACT FROM PUBLISHER]

We study the problem of identifying multiple rumor or infection sources in a network under the susceptible-infected model, and where these sources may start infection spreading at different times. We introduce the notion of an abstract estimator that, given the infection graph, assigns a higher value to each vertex in the graph it considers more likely to be a rumor source. This includes several of the single-source estimators developed in the literature. We introduce the concepts of a quasi-regular tree and a heavy center, which allows us to develop an algorithmic framework that transforms an abstract estimator into a two-source joint estimator, in which the infection graph can be thought of as covered by overlapping infection regions. We show that our algorithm converges to a local optimum of the estimation function if the underlying network is a quasi-regular tree. We further extend our algorithm to more than two sources, and heuristically to general graphs. Simulation results on both synthetic and real-world networks suggest that our algorithmic framework outperforms several existing multiple-source estimators, which typically assume that all sources start infection spreading at the same time. [ABSTRACT FROM PUBLISHER]

The product of the scanned responses of two colinear interleaved arrays estimates the spatial power spectral density of the received signal. This estimate has mathematical structure similar to the classical periodogram and hence is called the generalized periodogram. The generalized periodogram includes sparse arrays such as coprime arrays and two level nested arrays. This paper derives the analytical expected value of the generalized periodogram output for any signal. The generalized periodogram expected value is the convolution of the true spectral density with a windowing function. This result reveals that the expected product spectrum is not guaranteed to be positive definite. The paper also derives the covariance function of the generalized periodogram for a white Gaussian process. Finally, the paper illustrates the implicit equivalence between the product spectrum and estimation of the correlation function from the products of appropriately spaced pairs of sensors. [ABSTRACT FROM PUBLISHER]

This paper considers the problem of compressively sampling wide sense stationary random vectors with a low rank Toeplitz covariance matrix. Certain families of structured deterministic samplers are shown to efficiently compress a high-dimensional Toeplitz matrix of size N\times N , producing a compressed sketch of size O(\sqrt{r})\times O(\sqrt{r}).The reconstruction problem can be cast as that of line spectrum estimation, whereby, in absence of noise, Toeplitz matrices of any size N can be exactly recovered from compressive sketches of size O(\sqrtr)\times O(\sqrtr), no matter how large N is. In the presence of noise and finite data, the line spectrum estimation algorithm is combined with a novel denoising technique that only exploits a positive semidefinite (PSD) Toeplitz constraint to denoise the compressed sketch using a simple least-squares minimization framework. A major advantage of the algorithm is that it does not require any regularization parameter. It also enjoys lower computational complexity owing to its ability to predict the unobserved entries of the low-rank Toeplitz matrix. Explicit bounds on the reconstruction error are established and it is shown that the PSD constraint on the denoiser is sufficient to ensure stable reconstruction from a sketch of size O(\sqrtr)\times O(\sqrtr). Extensive simulations demonstrate that the proposed algorithm provides better performance over random samplers and algorithms that use nuclear-norm-based regularizers. [ABSTRACT FROM PUBLISHER]

In this paper, we consider an OFDM radio link corrupted by oscillator phase noise in the receiver, namely the problem of estimating and compensating for the impairment. To lessen the computational burden and delay incurred onto the receiver, we estimate phase noise using only scattered pilot subcarriers, i.e., no tentative symbol decisions are used in obtaining and improving the phase noise estimate. In particular, the phase noise estimation problem is posed as an unconstrained optimization problem whose minimizer suffers from the so-called amplitude and phase estimation error. These errors arise due to receiver noise, estimation from limited scattered pilot subcarriers and estimation using a dimensionality reduction model. It is empirically shown that, at high signal-to-noise-ratios, the phase estimation error is small. To reduce the amplitude estimation error, we restrict the minimizer to be drawn from the so-called phase noise geometry set when minimizing the cost function. The resulting optimization problem is a nonconvex program. However, using the S-procedure for quadratic equalities, we show that the optimal solution can be obtained by solving the convex dual problem. We also consider a less complex heuristic scheme that achieves the same objective of restricting the minimizer to the phase noise geometry set. Through simulations, we demonstrate improved coded bit-error-rate and phase noise estimation error performance when enforcing the phase noise geometry. For example, at high signal-to-noise-ratios, the probability density function of the phase noise estimation error exhibits thinner tails which results in lower bit-error-rate. [ABSTRACT FROM PUBLISHER]

We consider the problem of joint estimation of inverse covariance matrices lying in an unknown subspace of the linear space of symmetric matrices. We perform the estimation using groups of measurements with different covariances. Assuming the inverse covariances span a low-dimensional subspace, our aim is to determine this subspace and to exploit this knowledge in order to improve the estimation. We develop a novel optimization algorithm discovering and exploiting the underlying low-dimensional subspace. We provide a computationally efficient algorithm and derive a tight upper performance bound. Numerical simulations on synthetic and real world data are presented to illustrate the performance benefits of the algorithm. [ABSTRACT FROM PUBLISHER]

Spatial smoothing is a widely used preprocessing scheme to improve the performance of high-resolution parameter estimation algorithms in case of coherent signals or if only a small number of snapshots is available. In this paper, we present a first-order performance analysis of the spatially smoothed versions of $R$-D Standard ESPRIT and $R$-D Unitary ESPRIT for sources with arbitrary signal constellations as well as $R$-D NC Standard ESPRIT and $R$-D NC Unitary ESPRIT for strictly second-order (SO) non-circular (NC) sources. The derived expressions are asymptotic in the effective signal-to-noise ratio (SNR), i.e., the approximations become exact for either high SNRs or a large sample size. Moreover, no assumptions on the noise statistics are required apart from a zero mean and finite SO moments. We show that both $R$-D NC ESPRIT-type algorithms with spatial smoothing perform asymptotically identical in the high effective SNR regime. Generally, the performance of spatial smoothing based algorithms depends on the number of subarrays, which is a design parameter that needs to be chosen beforehand. In order to gain more insights into the optimal choice of the number of subarrays, we simplify the derived analytical $R$-D mean square error (MSE) expressions for the special case of a single source. The obtained MSE expression explicitly depends on the number of subarrays in each dimension, which allows us to analytically find the optimal number of subarrays for spatial smoothing. Based on this result, we additionally derive the maximum asymptotic gain from spatial smoothing and compute the asymptotic efficiency for the single source case in closed-form. All the analytical results are verified by simulations. [ABSTRACT FROM PUBLISHER]

One of the main objectives of nuclear spectroscopy is the estimation of the counting rate of unknown radioactive sources. Recently, we proposed an algorithm based on a sparse reconstruction of the time signal in order to estimate precisely this counting rate, under the assumption that it remained constant over time. Computable bounds were obtained to quantify the performances. This approach, based on a postprocessed approach of a non-negative sparse regression of the time signal, performed well even when the activity of the source was high. The purpose of this paper is to present an extension of the previous method for an activity varying over time. It relies on the same preliminary sparse reconstruction. However, the postprocessed and plug-in steps are made differently to fit the nonhomogeneous framework. The adapted bounds are presented, and results on simulations illustrate the advantages and limitations of this method. [ABSTRACT FROM PUBLISHER]

The problem of determining the intrinsic quality of a signal processing system with respect to the inference of an unknown deterministic parameter $\theta$ is considered. While the Fisher information measure $F(\theta)$ forms a classical tool for such a problem, direct computation of the information measure can become difficult in various situations. For the estimation theoretic performance analysis of nonlinear measurement systems, the form of the likelihood function can make the calculation of the information measure $F(\theta)$ challenging. In situations where no closed-form expression of the statistical system model is available, the analytical derivation of $F(\theta)$ is not possible at all. Based on the Cauchy–Schwarz inequality, we derive an alternative information measure $S(\theta)$. It provides a lower bound on the Fisher information $F(\theta)$ and has the property of being evaluated with the mean, the variance, the skewness, and the kurtosis of the system model at hand. These entities usually exhibit good mathematical tractability or can be determined at low-complexity by real-world measurements in a calibrated setup. With various examples, we show that $S(\theta)$ provides a good conservative approximation for $F(\theta)$ and outline different estimation theoretic problems where the presented information bound turns out to be useful. [ABSTRACT FROM PUBLISHER]

This paper studies a Dantzig-selector type regularized estimator for linear functionals of high-dimensional linear processes. Explicit rates of convergence of the proposed estimator are obtained and they cover the broad regime from independent identically distributed samples to long-range dependent time series and from sub-Gaussian innovations to those with mild polynomial moments. It is shown that the convergence rates depend on the degree of temporal dependence and the moment conditions of the underlying linear processes. The Dantzig-selector estimator is applied to the sparse Markowitz portfolio allocation and the optimal linear prediction for time series, in which the ratio consistency when compared with an oracle estimator is established. The effect of dependence and innovation moment conditions is further illustrated in the simulation study. Finally, the regularized estimator is applied to classify the cognitive states on a real functional magnetic resonance imaging dataset and to portfolio optimization on a financial dataset. [ABSTRACT FROM PUBLISHER]

As communication systems scale up in bandwidth, the limited resolution in high-speed analog-to-digital converters (ADCs) is a key challenge in realizing low-cost “mostly digital” transceiver architectures. This motivates a systematic effort to understand the limits of such architectures under the severe quantization constraints imposed by the use of low-precision ADCs. In particular, we investigate a canonical problem of blind carrier phase and frequency synchronization with coarse phase quantization in this paper. We develop a Bayesian approach to blind phase estimation, jointly modeling the unknown data, unknown phase and the quantization nonlinearity. We highlight the crucial role of dither, implemented via a mixed signal architecture with a digitally controlled phase shift prior to the ADC. We show the efficacy of random dither, and then improve upon its performance with a simple feedback control policy that is close to optimal in terms of rapidly reducing the mean squared error of phase estimation. This initial blind phase acquisition stage is followed by feedback-based phase/frequency tracking using an Extended Kalman Filter. Performance evaluations for a QPSK system show that excellent bit error rate (BER) performance, close to that of an unquantized system, is achieved by the use of 8 phase bins (implementable using 4 one-bit ADCs operating on linear combinations of in-phase and quadrature components). [ABSTRACT FROM PUBLISHER]

Recently, several high-resolution parameter estimation algorithms have been developed to exploit the structure of strictly second-order (SO) non-circular (NC) signals. They achieve a higher estimation accuracy and can resolve up to twice as many signal sources compared to the traditional methods for arbitrary signals. As a benchmark for these NC methods, we derive the closed-form deterministic $R$ -D NC Cramér-Rao bound (NC CRB) for the multi-dimensional parameter estimation of strictly non-circular (rectilinear) signal sources in this paper. Assuming a separable centro-symmetric $R$-D array, we show that in some special cases, the deterministic $R$-D NC CRB reduces to the existing deterministic $R$-D CRB for arbitrary signals. This suggests that no gain from strictly non-circular sources (NC gain) can be achieved under the deterministic data assumption in these cases. For more general scenarios, finding an analytical expression of the NC gain for an arbitrary number of sources is very challenging. Thus, in this paper, we simplify the derived NC CRB and the existing CRB for the special case of two closely-spaced strictly non-circular sources captured by a uniform linear array (ULA). Subsequently, we use these simplified CRB expressions to analytically compute the maximum achievable asymptotic NC gain for the considered two source case. The resulting expression only depends on the various physical parameters and we find the conditions that provide the largest NC gain. Our analysis is supported by extensive simulation results. [ABSTRACT FROM PUBLISHER]

In array processing, mutual coupling between sensors has an adverse effect on the estimation of parameters (e.g., DOA). Sparse arrays such as nested arrays, coprime arrays, and minimum redundancy arrays (MRA) have reduced mutual coupling compared to uniform linear arrays (ULAs). These arrays also have a difference coarray with O\left(N^2\right) virtual elements, where $N$ is the number of physical sensors, and can therefore resolve O\left(N^2\right) uncorrelated source directions. But these well-known sparse arrays have disadvantages: MRAs do not have simple closed-form expressions for the array geometry; coprime arrays have holes in the coarray; and nested arrays contain a dense ULA in the physical array, resulting in significantly higher mutual coupling than coprime arrays and MRAs. In a companion paper, a sparse array configuration called the (second-order) super nested array was introduced, which has many of the advantages of these sparse arrays, while removing most of the disadvantages. Namely, the sensor locations are readily computed for any $N$ (unlike MRAs), and the difference coarray is exactly that of a nested array, and therefore hole-free. At the same time, the mutual coupling is reduced significantly (unlike nested arrays). In this paper, a generalization of super nested arrays is introduced, called the $Q$th-order super nested array. This has all the properties of the second-order super nested array with the additional advantage that mutual coupling effects are further reduced for $Q > 2$. Many theoretical properties are proved and simulations are included to demonstrate the superior performance of these arrays. [ABSTRACT FROM PUBLISHER]

A maximum likelihood estimator for fusing the measurements in an inertial sensor array is presented. The maximum likelihood estimator is concentrated and an iterative solution method is presented for the resulting low-dimensional optimization problem. The Cramér–Rao bound for the corresponding measurement fusion problem is derived and used to assess the performance of the proposed method, as well as to analyze how the geometry of the array and sensor errors affect the accuracy of the measurement fusion. The angular velocity information gained from the accelerometers in the array is shown to be proportional to the square of the array dimension and to the square of the angular speed. In our simulations the proposed fusion method attains the Cramér–Rao bound and outperforms the current state-of-the-art method for measurement fusion in accelerometer arrays. Further, in contrast to the state-of-the-art method that requires a 3D array to work, the proposed method also works for 2D arrays. The theoretical findings are compared to results from real-world experiments with an in-house developed array that consists of 192 sensing elements. [ABSTRACT FROM PUBLISHER]