Statistical Analysis with Missing Data

System Identification I

Statistical Analysis With Missing Data (2nd ed.) (Book)

Fast Fourier Transform

Probability and Measure

Estimation of nonparametric noise and FRF models for multivariable systems—Part I: Theory

It is the common conviction that frequency domain system identification suffers from the drawback that it cannot handle arbitrary signals without introducing systematic errors. This paper shows that it is possible to deal with nonperiodic signals without any approximation and under the same assumptions as in the time domain, by estimating simultaneously some initial conditions and the system model parameters.

Frequency domain system identification using arbitrary signals

Frequency-domain system identification using non-parametric noise models estimated from a small number of data sets

This paper presents a study of possible models to describe the relation between the scintillation light point-of-origin and the measured photo detector pixel signals in monolithic scintillation crystals From these models the X, Y and depth of interaction (DOI) coordinates can be estimated simultaneously by nonlinear least-square fitting The method depends only on the information embedded in the signals of individual events, and therefore does not need any prior position training or calibration Three possible distributions of the light sources were evaluated: an exact solid-angle-based distribution, an approximate solid-angle distribution and an extended approximate solid-angle-based distribution which includes internal reflection at side and bottom surfaces The performance of the general model using these three distributions was studied using Monte Carlo simulated data of a 20 × 20 × 10 mm lutetium oxyorthosilicate (Lu2SiO5 or LSO) block read out by 2 Hamamatsu S8550 avalanche photo diode arrays The approximate solid-angle-based model had the best compromise between resolution and simplicity This model was also evaluated using experimental data by positioning a narrow 12 mm full width at half maximum (FWHM) beam of 511 keV photons at known positions on the 20 × 20 × 10 mm LSO block An average intrinsic resolution in the X-direction of 14 mm FWHM was obtained for positions covering the complete block The intrinsic DOI resolution was estimated at 26 mm FWHM

http://iopscience.iop.org/article/10.1088/0031-9155/55/21/012/pdf

Nonlinear least-squares modeling of 3D interaction position in a monolithic scintillator block.

The worldwide available frequency bands around 60GHz offer new opportunities for high-data-rate communication. Despite the large amount of research on CMOS mm-wave frequency generation, current state-of-the-art mm-wave PLLs [1–5] do not achieve a tuning range that is wide enough to cover the worldwide union of bands around 60GHz (57 to 66GHz), with some margin for process and temperature variations. Achieving a large tuning range is hampered by the limited varactor tunability at 60GHz, and the need for a high-frequency prescaler with a wide locking range.

Gerd Vandersteen

Papers

Estimation of nonparametric noise and FRF models for multivariable systems—Part I: Theory

Frequency domain system identification using arbitrary signals

Frequency-domain system identification using non-parametric noise models estimated from a small number of data sets

Nonlinear least-squares modeling of 3D interaction position in a monolithic scintillator block.

A 57-to-66GHz quadrature PLL in 45nm digital CMOS