Distributed variational Bayesian algorithms for Gaussian mixtures in sensor networks

• Published in: Signal processing Vol. 90; no. 4; pp. 1197 - 1208
• Main Authors: Safarinejadian, B., Menhaj, M.B., Karrari, M.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.04.2010; Elsevier
• Subjects: Algorithms; Applied sciences; Bayesian analysis; Clustering; Density; Density estimation; Detection, estimation, filtering, equalization, prediction; Estimates; Exact sciences and technology; Gaussian; Information, signal and communications theory; Mathematical models; Mixture of Gaussians; Networks; Pattern recognition; Sensor networks; Sensors; Signal and communications theory; Signal processing; Signal representation. Spectral analysis; Signal, noise; Telecommunications and information theory; Variational approximations; Sensor networks; Density estimation; Variational approximations; Clustering; Mixture of Gaussians; Performance evaluation; Automatic classification; Variational methods; Mixture theory; Pattern recognition; Signal classification; Statistical method; Distributed processing; Gaussian process; Simulation; Pattern classification; Distributed algorithm; Variational principle; Localization; Sensor array; Environmental monitoring; Target detection; Signal detection
• ISSN: 0165-1684; 1872-7557
• DOI: 10.1016/j.sigpro.2009.10.004

This paper presents a distributed variational Bayesian (DVBA) algorithm for density estimation and clustering in sensor networks. It is assumed that measurements of the sensors can be statistically modeled by a common Gaussian mixture model. The variational approach allows the simultaneous estimate of the component parameters and the model complexity. The DVBA algorithm produces an estimate of the density of the sensor data without requiring the data to be transmitted to and processed at a central location. Alternatively, DVBA can be viewed as a distributed processing approach for clustering the sensor data into components corresponding to predominant environmental features sensed by the network. The convergence of the proposed DVBA is also investigated. The proposed method is then used for environmental monitoring and also distributed target classification. Simulation results approve promising performance of this algorithm.