Kalman Filtering With Unlimited Sensing

Hongwei Wang; Xi Zheng; Hongbin Li

ICASSP 2024accepted0 citations

Kalman Filtering With Unlimited Sensing

Abstract

In this paper, we consider state estimation in the Kalman filtering framework with unlimited sensing measurements (USMs), which are obtained from sensors equipped with a self-reset analog-to-digital (SR-ADC). SR-ADC was recently introduced to deal with the saturation issue frequently encountered in a conventional ADC. To tackle the nonlinearity of the USM, we present a unique decomposition property of the USM. Leveraging this property and a multiple model adaptive estimation strategy, we propose a novel USF-based Kalman filtering (KF-USM) algorithm. Numerical results reveal that the proposed KF-USM filter is an effective alternative to the conventional ADC-based KF to deal with high dynamic range input signals, offering more accurate state estimation in the presence of saturation.

BibTeX

@inproceedings{icassp2024_kalmanfilteringw,
  title = {Kalman Filtering With Unlimited Sensing},
  author = {Hongwei Wang and Xi Zheng and Hongbin Li},
  booktitle = {ICASSP 2024},
  year = {2024}
}