Adaptive detection schemes in compound-Gaussian clutter

Radar detection of coherent pulse trains embedded in compound-Gaussian disturbance with partially known statistics is discussed. We first give a thorough derivation of two recently proposed adaptive detection structures. Next, we derive a different detection scheme exploiting the assumption that the clutter is wide-sense stationary. Resorting to the theory of circulant matrices, in fact, we demonstrate that the estimation of the structure of the clutter covariance matrix can be reduced to the estimation of its eigenvalues, which in turn can be (efficiently) done via fast Fourier transform codes. After a thorough performance assessment, mostly carried on via computer simulations, the results show that the newly proposed detector achieves better performance than the two previously introduced adaptive detectors. Moreover, a sensitivity analysis shows that, even though this detector does not strictly guarantee the constant false alarm rate property with respect to the clutter covariance matrix, it is robust, in the sense that its performance is only slightly affected by variations in the clutter temporal correlation     

Fault Detection in Dynamic Systems via Decision Fusion

Due to the growing demands for system reliability and availability of large amounts of data, efficient fault detection techniques for dynamic systems are desired. In this paper, we consider fault detection in dynamic systems monitored by multiple sensors. Normal and faulty behaviors can be modeled as two hypotheses. Due to communication constraints, it is assumed that sensors can only send binary data to the fusion center. Under the assumption of independent and identically distributed (1ID) observations, we propose a distributed fault detection algorithm, including local detector design and decision fusion rule design, based on state estimation via particle filtering. Illustrative examples are presented to demonstrate the effectiveness of our approach.     

A novel resonant converter topology for DC-to-DC power supply

A class-E DC-to-DC converter with half-wave controlled current rectifier is proposed. Its output voltage is controlled by the conduction angle of the rectifier switch at constant switching frequency. Zero voltage switching for all the switches can be maintained from full load to no load. Its steady state characteristics are analyzed and the effects of the circuit parameters are studied. Some extensions of the proposed converter are also discussed. The analysis is verified by PSPICE simulation and an experimental prototype     

The THEMIS Array of Ground-based Observatories for the Study of Auroral Substorms

The NASA Time History of Events and Macroscale Interactions during Substorms (THEMIS) project is intended to investigate magnetospheric substorm phenomena, which are the manifestations of a basic instability of the magnetosphere and a dominant mechanism of plasma transport and explosive energy release. The major controversy in substorm science is the uncertainty as to whether the instability is initiated near the Earth, or in the more distant >20 Re magnetic tail. THEMIS will discriminate between the two possibilities by using five in-situ satellites and ground-based all-sky imagers and magnetometers, and inferring the propagation direction by timing the observation of the substorm initiation at multiple locations in the magnetosphere. An array of stations, consisting of 20 all-sky imagers (ASIs) and 30-plus magnetometers, has been developed and deployed in the North American continent, from Alaska to Labrador, for the broad coverage of the nightside magnetosphere. Each ground-based observatory (GBO) contains a white light imager that takes auroral images at a 3-second repetition rate (“cadence”) and a magnetometer that records the 3 axis variation of the magnetic field at 2 Hz frequency. The stations return compressed images, “thumbnails,” to two central databases: one located at UC Berkeley and the other at the University of Calgary, Canada. The full images are recorded at each station on hard drives, and these devices are physically returned to the two data centers for data copying. All data are made available for public use by scientists in “browse products,” accessible by using internet browsers or in the form of downloadable CDF data files (the “browse products” are described in detail in a later section). Twenty all-sky imager stations are installed and running at the time of this publication. An example of a substorm was observed on the 23rd of December 2006, and from the THEMIS GBO data, we found that the substorm onset brightening of the equatorward arc was a gradual process (>27 seconds), with minimal morphology changes until the arc breaks up. The breakup was timed to the nearest frame (<3 s) and located to the nearest latitude degree at about ±3^oE in longitude. The data also showed that a similar breakup occurred in Alaska ～10 minutes later, highlighting the need for an array to distinguish prime onset.     

Almost instant time inference for hybrid partially dynamic Bayesian networks

A Bayesian network (BN) is a compact representation for probabilistic models and inference. They have been used successfully for many military and civilian applications. It is well known that, in general, the inference algorithms to compute the exact a posterior probability of a target node given observed evidence are either computationally infeasible for dense networks or impossible for general hybrid networks. In those cases, one either computes the approximate results using stochastic simulation methods or approximates the model using discretization or a Gaussian mixture model before applying an exact inference algorithm. This paper combines the concept of simulation and model approximation to propose an efficient algorithm for those cases. The main contribution here is a unified treatment of arbitrary (nonlinear non-Gaussian) hybrid (discrete and continuous) BN inference having both computation and accuracy scalability. The key idea is to precompile the high-dimensional hybrid distribution using a hypercube representation and apply it for both static and dynamic BN inference. Since the inference process essentially becomes a combination of table look-up and some simple operations, the method is shown to be extremely efficient. It can also he scaled to achieve any desirable accuracy given sufficient preprocessing time and memory for the cases where exact inference is not possible     

Covariance matrix estimation for adaptive CFAR detection incompound-Gaussian clutter

We address the estimation of the structure of the covariance matrix and its application to adaptive radar detection of coherent pulse trains in clutter-dominated disturbance modeled as a compound-Gaussian process. For estimation purposes we resort to range cells in spatial proximity with that under test and assume that these cells, free of signal components, can be clustered into groups of data with one and the same value of the texture. We prove that, plugging the proposed estimator of the structure of the covariance matrix into a previously derived detector, based upon the generalized likelihood ratio test (GLRT), leads to an adaptive detector which ensures the constant false alarm rate (CFAR) property with respect to the clutter covariance matrix as well as the statistics of the texture. Finally, we show that this adaptive receiver has an acceptable loss with respect to its nonadaptive counterpart in cases of relevant interest for radar applications     

The relation of Algols and W Serpentis stars

Evidence on the issues of whether the W Serpentis stars are a coherent class, and how they may interface with the Algol systems, is reviewed, with emphasis on the idea that they are semi-detached systems in the latter part of the rapid phase of mass transfer, with optically and geometrically thick disks of transferred gas around the (now) more massive star. We are interested in what will be seen when the gas clears away, and mainly examine the idea that it will be an Algol-type system. More particularly, consideration is given to centrifugally limited accretion as a mechanism to build up a substantial disk, and the presumed evolutionary sequence is from a W Ser to a rapidly rotating Algol to a normal Algol system. Systems such as V367 Cyg and RW Tau fit into this scheme only with difficulty. Because it is extremely difficult to measure the rotation of some W Ser (mass) primaries, it is natural to look at the rotation statistics of Algols to test this idea. The badly behaved light curves and spectroscopy of some Algols (eg. U Cep, RZ Sct) may be attributable to the double contact condition, and the ramifications of this possibility are discussed. If so, the rotation statistics of Algols should show two spikes, corresponding to the two special conditions into which a system should be driven by tidal braking and centrifugally limited spinup. Present rotation statistics do show these spikes. Algols should flip between these states fairly quickly, depending on the mass transfer rate. Thus, to the extent that the meager statistics can be accepted as meaningful, the new (fourth) morphological type of close binary (double contact) has attained demonstrable reality. The rotation statistics are presented in terms of a particular rotation parameter, R, which is zero for synchronism and unity for the centrifugal limit. Future work should develop rotation statistics to see if the rotational lobe-filling (R = 1) spike persists. It should also look into whether W Ser primaries are on the hydrogen burning main sequence, or in general what they are. We also need more light curves of W Ser type systems, high resolution line profiles for the (mass) primaries (with particular attention to the W Ser-Algol transition cases), and spectroscopy of low inclination W Serpentis systems, such as KX And.     

Analytical solution for continuous-time Kalman tracking filterswith colored measurement noise in frequency domain

New expressions are given for analytical solutions to the steady-state Kalman gains of the two-state exponentially correlated velocity (ECV) and the three-state exponentially correlated acceleration (ECA) tracking filters with position measurements by using spectral factorization method. The measurement colored noise model is characterized by a correlation time 1/λ. The vehicle oscillations such as wind-induced-bending is also considered in the modeling of the system which leads to the most generalized state transition matrix