In this paper we propose a solution to the problem of tracking quasi-periodically varying systems based on the local basis function (LBF) approach. Within this framework, parameter trajectories are locally approximated using linear combinations of specific functions of time known as basis functions. We derive both bias and variance characteristics of LBF estimators. Additionally, we demonstrate that the computational burden associated with LBF estimation algorithms can be significantly reduced, without sacrificing high estimation accuracy, by employing the computationally fast, approximate version of the LBF scheme.
Authors
Additional information
- DOI
- Digital Object Identifier link open in new tab 10.23919/eusipco63174.2024.10715241
- Category
- Aktywność konferencyjna
- Type
- publikacja w wydawnictwie zbiorowym recenzowanym (także w materiałach konferencyjnych)
- Language
- angielski
- Publication year
- 2024
