<![CDATA[ IEEE Transactions on Signal Processing - new TOC ]]>
http://ieeexplore.ieee.org
TOC Alert for Publication# 78 2018March 19<![CDATA[Editorial A Brief Message From the New Editor-in-Chief]]>6681952195276<![CDATA[Noise Enhancement in Robust Estimation of Location]]>66819531966839<![CDATA[Large-Scale Kernel-Based Feature Extraction via Low-Rank Subspace Tracking on a Budget]]>nonparametric nature of kernel-based estimators are computational and memory requirements that become prohibitive with large-scale datasets. In response to this formidable challenge, this paper puts forward a low-rank, kernel-based, feature extraction approach that is particularly tailored for online operation. A novel generative model is introduced to approximate high-dimensional (possibly infinite) features via a low-rank nonlinear subspace, the learning of which lends itself to a kernel function approximation. Offline and online solvers are developed for the subspace learning task, along with affordable versions, in which the number of stored data vectors is confined to a predefined budget. Analytical results provide performance bounds on how well the kernel matrix as well as kernel-based classification and regression tasks can be approximated by leveraging budgeted online subspace learning and feature extraction schemes. Tests on synthetic and real datasets demonstrate and benchmark the efficiency of the proposed method for dynamic nonlinear subspace tracking as well as online classification and regressions tasks.]]>668196719812008<![CDATA[Optimal Bayesian Kalman Filtering With Prior Update]]>posterior effective noise statistics , which are found by employing the method of factor graphs through formulating the problem of computing the likelihood function as a message passing algorithm.]]>668198219961877<![CDATA[Subspace Rejection for Matching Pursuit in the Presence of Unresolved Targets]]>668199720101178<![CDATA[Relay Hybrid Precoding Design in Millimeter-Wave Massive MIMO Systems]]>668201120261606<![CDATA[Asymmetric Pulse Modeling for FRI Sampling]]>unique for parametrically modeling pulse asymmetry. The FrH operator is obtained by a trigonometric interpolation between the standard Hilbert and identity operators, where the interpolation weights are determined by the degree of asymmetry. The FrH operators are also steerable, which allows for estimation of the asymmetry factors, in addition to the delays and amplitudes, using the high-resolution spectral estimation techniques that are used for solving standard FRI problems. We also develop the discrete counterpart using discrete FrH operators and show that all the desirable properties carry over smoothly to the discrete setting as well. We derive closed-form expressions for the Cramér–Rao bounds and Hammersley–Chapman–Robbins bound, on the variances of the estimators for continuous and discrete parameters, respectively. Experimental results show that the proposed estimators have variances that meet the lower bounds. We demonstrate an application of the proposed discrete FrH methodology on real electrocardiogram (ECG) signals in the presence of noise. Specifically, we show how the asymmetry of QRS complexes in various channels of an ECG signal could be modeled accurately.]]>668202720401763<![CDATA[Riemannian Optimization and Approximate Joint Diagonalization for Blind Source Separation]]>668204120541539<![CDATA[Blind Radio Tomography]]>668205520691958<![CDATA[Attack Detection in Sensor Network Target Localization Systems With Quantized Data]]>668207020851092<![CDATA[Uniform Recovery Bounds for Structured Random Matrices in Corrupted Compressed Sensing]]>$s$ -sparse signal $mathbf{x}^{star }in mathbb {C}^n$ from corrupted measurements $mathbf{y}=mathbf{A}mathbf{x}^{star }+mathbf{z}^{star }+mathbf{w}$, where $mathbf{z}^{star }in mathbb {C}^m$ is a $k$-sparse corruption vector whose nonzero entries may be arbitrarily large and $mathbf{w}in mathbb {C}^m$ is a dense noise with bounded energy. The aim is to exactly and stably recover the sparse signal with tractable optimization programs. In this paper, we prove the uniform recovery guarantee of this problem for two classes of structured sensing matrices. The first class can be expressed as the product of a unit-norm tight frame (UTF), a random diagonal matrix, and a bounded columnwise orthonormal matrix (e.g., partial random circulant matrix). When the UTF is bounded (i.e. $mu (mathbf{U})sim 1/sqrt{m}$), we prove that with high probability, one can recover an $s$-sparse signal exactly and stably by $l_1$ minimization programs even if the measurements are corrupted by a sparse vector, provided $m=mathcal{O}(slog ^2slog ^2n)$ and the sparsity level $k$ of the corruption is a constant fraction of the total number of measur-
ments. The second class considers a randomly subsampled orthonormal matrix (e.g., random Fourier matrix). We prove the uniform recovery guarantee provided that the corruption is sparse on certain sparsifying domain. Numerous simulation results are also presented to verify and complement the theoretical results.]]>66820862097523<![CDATA[Optimized Update/Prediction Assignment for Lifting Transforms on Graphs]]>$mathcal{U}$) and prediction ($mathcal{P}$) nodes. This is the update/prediction ($mathcal{U}/mathcal{P}$) assignment problem, which is the focus of this paper. We analyze this problem theoretically and derive an optimal $mathcal{U}/mathcal{P}$ assignment under assumptions about signal model and filters. Furthermore, we prove that the best $mathcal{U}/mathcal{P}$ partition is related to the correlation between nodes on the graph and is not the one that minimizes the number of conflicts (connections between nodes of same label) or maximizes the weight of the cut. We also provide experimental results in randomly generated graph signals and real data from image and video signals that validate our theoretical conclusions, demonstrating improved performance over state-of-the-art solutions for this problem.]]>668209821113103<![CDATA[Multiple Scan Data Association by Convex Variational Inference]]>668211221271606<![CDATA[On the Use of Extrinsic Probabilities in the Computation of Non-Bayesian Cramér–Rao Bounds for Coded Linearly Modulated Signals]]>ad hoc CRB (ACRB) expressions; the first ACRB is obtained by substituting in the exact CRB expression for uncoded modulation the a priori symbol probabilities by the extrinsic symbol probabilities; the second ACRB, which has received some attention in recent scientific publications, additionally assumes that the real and imaginary parts of the symbols are independent. Our exposition focuses on the particular case of phase shift estimation. By means of examples we show that although for some coded modulation schemes the exact and ACRBs yield virtually the same numerical result, for other coded modulation schemes the ACRBs differ considerably among themselves and from the exact CRB. We provide some explanations for this behavior. We also argue that both ACRBs are expected to virtually coincide with the exact CRB in the case of bit-interleaved coded modulation combined with a rectangular constellations with independent in-phase and quadrature mapping and a binary code for which the factor graph does not contain short cycles.]]>668212821401266<![CDATA[ToPs: Ensemble Learning With Trees of Predictors]]>tree of predictors. The (locally) optimal tree of predictors is derived recursively; each step involves jointly optimizing the split of the terminal nodes of the previous tree and the choice of learner (from among a given set of base learners) and training set—hence predictor—for each set in the split. The features of a new instance determine a unique path through the optimal tree of predictors; the final prediction aggregates the predictions of the predictors along this path. Thus, our approach uses base learners to create complex learners that are matched to the characteristics of the data set while avoiding overfitting. We establish loss bounds for the final predictor in terms of the Rademacher complexity of the base learners. We report the results of a number of experiments on a variety of datasets, showing that our approach provides statistically significant improvements over a wide variety of state-of-the-art machine learning algorithms, including various ensemble learning methods.]]>668214121521772<![CDATA[Structure-Aware Bayesian Compressive Sensing for Frequency-Hopping Spectrum Estimation With Missing Observations]]>668215321661041<![CDATA[A Robust Parallel Algorithm for Combinatorial Compressed Sensing]]>$mathbf{x} in mathbb {R}^n$ with at most $k < n$ nonzeros can be recovered from an expander sketch $mathbf{A}mathbf{x}$ in $mathcal{O}(text{nnz}(mathbf{A})log k)$ operations via the parallel-$ell _0$ decoding algorithm, where $text{nnz}(mathbf{A})$ denotes the number of nonzero entries in $mathbf{A} in mathbb {R}^{m times n}$. In this paper, we present the robust-$ell _0$ decoding algorithm, which robustifies parallel-$ell _0$ when the sketch $mathbf{A}mathbf{x}$ is corrupted by additive noise. This robustness is achieved by approximating the asymptotic posterior distribution of values in the sketch given its corrupted measurements. We provide analytic expressions that approximate these posteriors under the assumptions that the nonzero entries in the signal and the noise are drawn from continuous distributions. Numerical experiments presented show that robust- $ell _0$ is superior to existing greedy and combinatorial compressed sensing algorithms in the presence of small to moderate signal-to-noise ratios in the setting of Gaussian signals and Gaussian additive noise.]]>668216721771257<![CDATA[Asymptotic Confidence Regions for High-Dimensional Structured Sparsity]]>et al., then using an appropriate estimator for the precision matrix $Theta$. In order to estimate the precision matrix a corresponding structured matrix norm penalty has to be introduced. After normalization the result is an asymptotic pivot. The asymptotic behavior is studied and simulations are added to study the differences between the two schemes.]]>668217821901440<![CDATA[Navigation With Cellular CDMA Signals—Part I: Signal Modeling and Software-Defined Receiver Design]]>668219122035991<![CDATA[Navigation With Cellular CDMA Signals—Part II: Performance Analysis and Experimental Results]]>668220422184218