Search

This paper studies a statistical model for heteroscedastic (i.e., power fluctuating) signals embedded in white Gaussian noise. Using the Riemannian geometry theory, we propose an unified approach to tackle several problems related to this model. The first axis of contribution concerns parameters (signal subspace and power factors) estimation, for which we derive intrinsic Cramér-Rao bounds and propose a flexible Riemannian optimization algorithmic framework in order to compute the maximum likelihood estimator (as well as other cost functions involving the parameters). Interestingly, the obtained bounds are in closed forms and interpretable in terms of problem’s dimensions and SNR. The second axis of contribution concerns the problem of clustering data assuming a mixture of heteroscedastic signals model, for which we generalize the Euclidean K-means++ to the considered Riemannian parameter space. We propose an application of the resulting clustering algorithm on the Indian Pines segmentation problem benchmark. [ABSTRACT FROM AUTHOR]

Radio sources in the Very High Frequency (VHF) band can be seized as opportunity donors in a passive radar configuration such as FM radio stations and VHF omnidirectional range (VOR). A full‐wave simulation of three size classes of aeroplanes shows that their bistatic radar cross‐section (RCS) are statistically comparable, albeit perform differently in time while the plane is flying. This difference can be exploited to recognize the size of the aeroplanes with respect to these classes. Measurements confirm this possible differentiation between the aeroplanes within the same class. Encouraging initial results were obtained using convolutional or recurrent neural networks to classify aircraft classes, combining simulated bistatic RCS results and real trajectories (collected from automatic dependent surveillance‐broadcast data). [ABSTRACT FROM AUTHOR]

The estimation of covariance matrices is a core problem in many modern adaptive signal processing applications. For matrix- and array-valued data, e.g., MIMO communication, EEG/MEG (time versus channel), the covariance matrix of vectorized data may belong to the non-convex set of Kronecker product structure. In addition, the Kronecker factors can also exhibit an additional linear structure. Taking this prior knowledge into account during the estimation process drastically reduces the amount of unknown parameters, and then improves the estimation accuracy. On the other hand, the broad class of complex elliptically symmetric distributions, as well as the related complex angular elliptical distribution, are particularly suited to model heavy-tailed multivariate data. In this context, we first establish novel robust estimators of scatter and shape matrices (both related to a covariance matrix), having a Kronecker product structure with linearly structured Kronecker factors. Then, we conduct a theoretical analysis of their asymptotic performance (i.e., consistency, asymptotic distribution and efficiency), in matched and mismatched scenarios, i.e., when misspecifications between the true and assumed models occur. Finally, numerical results illustrate the theoretical analysis and assess the usefulness of the proposed estimators. [ABSTRACT FROM AUTHOR]

The joint estimation of means and scatter matrices is often a core problem in multivariate analysis. In order to overcome robustness issues, such as outliers from Gaussian assumption, $M$ -estimators are now preferred to the traditional sample mean and sample covariance matrix. These estimators are well established and studied in the real case since the seventies. Their extension to the complex case has drawn recent interest. In this letter, we derive the asymptotic performance of complex $M$ -estimators for multivariate location and scatter matrix estimation. [ABSTRACT FROM AUTHOR]

In this letter, we study the behavior of the maximum likelihood estimator (MLE) in the framework of low noise level (or high signal-to-noise ratio), when the data follow a unimodal elliptical distribution. The MLE appears to be the same as in the Gaussian context, regardless the noise distribution. We also show that the asymptotic distribution of this estimator is unimodal elliptical, where the law is intimately linked to that of the noise distribution. Additionally, this estimator is shown to be not efficient, except in the Gaussian noise case. Finally, we validate our analytic results by some simulations. [ABSTRACT FROM AUTHOR]

In this letter, we investigated the connection between information and estimation measures for mismatched Gaussian models. In addition to the input prior mismatch, we take into account the noise mismatch and establish a new relation between relative entropy and excess mean square error. The derived formula shows that the input prior mismatch may be canceled by the noise mismatch. Finally, an example illustrates the impact of model mismatches on estimation accuracy. [ABSTRACT FROM PUBLISHER]

This letter investigates hybrid lower bounds on the mean square error in order to predict the so-called threshold effect. A new family of tighter hybrid large error bounds based on linear transformations (discrete or integral) of a mixture of the McAulay–Seidman bound and the Weiss–Weinstein bound is provided in multivariate parameters case with multiple test points. For use in applications, we give a closed-form expression of the proposed bound for a set of Gaussian observation models with parameterized mean, including tones estimation which exemplifies the threshold prediction capability of the proposed bound. [ABSTRACT FROM PUBLISHER]