Gaussian mixture parameter estimation for cognitive radio and network surveillance applications

In cognitive radio networks a secondary user needs to estimate the primary users' air traffic patterns so as to optimize its transmission strategy. In this chapter, the authors describe a nonparametric Bayesian method for identifying traffic applications, since the traffic applications have their own distinctive air traffic patterns. In the proposed algorithm, the collapsed Gibbs sampler is applied to cluster the air traffic applications using the infinite Gaussian mixture model over the feature space of the packet length, the packet inter-arrival time, and the variance of packet lengths. The authors analyze the effectiveness of their proposed technique by extensive simulation using the measured data obtained from the WiMax networks.

Download Full-text

Rainfall-Rate Estimation Using Gaussian Mixture Parameter Estimator: Training and Validation

Journal of Atmospheric and Oceanic Technology ◽

10.1175/jtech-d-11-00122.1 ◽

2012 ◽

Vol 29 (5) ◽

pp. 731-744 ◽

Cited By ~ 3

Author(s):

Zhengzheng Li ◽

Yan Zhang ◽

Scott E. Giangrande

Keyword(s):

General Framework ◽

Gaussian Mixture ◽

Minimum Variance ◽

Rainfall Rate ◽

Polarimetric Radar ◽

Parameter Estimator ◽

Minimum Variance Unbiased Estimator ◽

Rate Estimation ◽

Parameter Estimations ◽

Mixture Parameter

Abstract This study develops a Gaussian mixture rainfall-rate estimator (GMRE) for polarimetric radar-based rainfall-rate estimation, following a general framework based on the Gaussian mixture model and Bayes least squares estimation for weather radar–based parameter estimations. The advantages of GMRE are 1) it is a minimum variance unbiased estimator; 2) it is a general estimator applicable to different rain regimes in different regions; and 3) it is flexible and may incorporate/exclude different polarimetric radar variables as inputs. This paper also discusses training the GMRE and the sensitivity of performance to mixture number. A large radar and surface gauge observation dataset collected in central Oklahoma during the multiyear Joint Polarization Experiment (JPOLE) field campaign is used to evaluate the GMRE approach. Results indicate that the GMRE approach can outperform existing polarimetric rainfall techniques optimized for this JPOLE dataset in terms of bias and root-mean-square error.

Download Full-text

Regularized Parameter Estimation in High-Dimensional Gaussian Mixture Models

Neural Computation ◽

10.1162/neco_a_00128 ◽

2011 ◽

Vol 23 (6) ◽

pp. 1605-1622 ◽

Cited By ~ 12

Author(s):

Lingyan Ruan ◽

Ming Yuan ◽

Hui Zou

Keyword(s):

Parameter Estimation ◽

Mixture Models ◽

Gaussian Mixture Models ◽

Expectation Maximization Algorithm ◽

Real Data ◽

Gaussian Mixture ◽

High Dimensional ◽

Model Based Clustering ◽

Text Type ◽

Effective Dimensionality

Finite gaussian mixture models are widely used in statistics thanks to their great flexibility. However, parameter estimation for gaussian mixture models with high dimensionality can be challenging because of the large number of parameters that need to be estimated. In this letter, we propose a penalized likelihood estimator to address this difficulty. The [Formula: see text]-type penalty we impose on the inverse covariance matrices encourages sparsity on its entries and therefore helps to reduce the effective dimensionality of the problem. We show that the proposed estimate can be efficiently computed using an expectation-maximization algorithm. To illustrate the practical merits of the proposed method, we consider its applications in model-based clustering and mixture discriminant analysis. Numerical experiments with both simulated and real data show that the new method is a valuable tool for high-dimensional data analysis.

Download Full-text