A hybrid reconstruction algorithm for computerized ionospheric tomography

Abstract

Computerized Ionospheric Tomography (CIT) is a method to reconstruct ionospheric electron density images by using the Global Positioning System data collected by the earth based receivers. In this study, Total Electron Content values obtained from a model based ionosphere and tomographic reconstruction techniques are used together to obtain ionospheric electron density distribution. Algebraic Reconstruction Technique (ART) is one of the most commonly used reconstruction method in medical tomography due to its simplicity in implementation. The performance of ART is independent of basis functions and very sensitive to the initial state. Total Least Squares (TLS) algorithm assumes no regularization and produces the lowest error for Haar basis for a given Latitude interval. The performance of TLS is improved with the number of receivers. If only one receiver is used, TLS algorithm together with Haar basis functions produces a low computational complexity and has a lower reconstruction error compared to Regularized Least Squares Algorithm, When the estimation by TLS is input as the initial state of ART, the overall reconstruction error reduces significantly compared to the reconstruction error of ART only or TLS with Haar basis only.