IEEE TSIPN Article

In this paper, we consider the problem of bandwidth-constrained distributed estimation of a Gaussian vector with linear observation model. Each sensor makes a scalar noisy observation of the unknown vector, employs a multi-bit scalar quantizer to quantize its observation, and maps it to a digitally modulated symbol.

Distributed machine learning algorithms enable learning of models from datasets that are distributed over a network without gathering the data at a centralized location. While efficient distributed algorithms have been developed under the assumption of faultless networks, failures that can render these algorithms nonfunctional occur frequently in the real world. 

In this paper, a multi-hypothesis distributed detection technique with non-identical local detectors is investigated. Here, for a global event, some of the sensors/detectors can observe the whole set of hypotheses, whereas the remaining sensors can either see only some aspects of the global event or infer more than one hypothesis as a single hypothesis.

We study the problem of distributed filtering for state space models over networks, which aims to collaboratively estimate the states by a network of nodes. Most of existing works on this problem assume that both process and measurement noises are Gaussian and their covariances are known in advance. In some cases, this assumption breaks down and no longer holds.

Expander recovery is an iterative algorithm designed to recover sparse signals measured with binary matrices with linear complexity. In the paper, we study the expander recovery performance of the bipartite graph with girth greater than 4, which can be associated with a binary matrix with column correlations equal to either 0 or 1. 

A key challenge in designing distributed particle filters is to minimize the communication overhead without compromising tracking performance. In this paper, we present two distributed particle filters that achieve robust performance with low communication overhead.

The adaptive algorithms applied to distributed networks are usually real-valued diffusion subband adaptive filter algorithms. However, it cannot be used for processing the complex-valued signals. In this paper, a novel augmented complex-valued diffusion normalized subband adaptive filter (D-ACNSAF) algorithm is proposed for distributed estimation over networks. In order to deal with the noncircular complex-valued signals, the D-ACNSAF algorithm uses the widely linear model for a diffusion network.

This paper utilizes the family of affine projection algorithms (APAs) for distributed estimation in the adaptive diffusion networks. The diffusion APA (DAPA), the diffusion selective partial update (SPU) APA (DSPU-APA), the diffusion selective regressor (SR) APA (DSR-APA), and the diffusion dynamic selection (DS) APA (DDS-APA) are introduced in a unified way. In DSPU-APA, the weight coefficients are partially updated at each node during the adaptation.

We consider the detection and tracking of a target in a decentralized sensor network. The presence of the target is uncertain, and the sensor measurements are affected by clutter and missed detections. The state-evolution model and the measurement model may be nonlinear and non-Gaussian. For this practically relevant scenario, we propose a particle-based distributed Bernoulli filter (BF) that provides to each sensor approximations of the Bayes-optimal estimates of the target presence probability and the target state.

This paper studies the problem of estimation from relative measurements in a graph, in which a vector indexed over the nodes has to be reconstructed from pairwise measurements of differences between its components associated with nodes connected by an edge. In order to model heterogeneity and uncertainty of the measurements, we assume them to be affected by additive noise distributed according to a Gaussian mixture.