AUTHOR=Weiss Taeer , Routtenberg Tirza , Ostrometzky Jonatan , Messer Hagit 

TITLE=Intensity estimation after detection for accumulated rainfall estimation

JOURNAL=Frontiers in Signal Processing

VOLUME=Volume 4 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/signal-processing/articles/10.3389/frsip.2024.1291878

DOI=10.3389/frsip.2024.1291878

ISSN=2673-8198

ABSTRACT=This work focuses on the optimization of the process of estimating the accumulated rain from measurements of the attenuation level of signals from commercial microwave links (CMLs). The process of accumulated rain estimation is usually based on estimation after detection, where first it is determined if there is rain at a specific time period, and then the accumulated rain at the detected rainy period is estimated. Naturally, errors in the detection step affect the accuracy of the consequent accumulated rain estimation. However, traditionally, the detection and the estimation steps are designed independently. In particular, the detection threshold in this setting, i.e., the lowest level that would be declared as rain, is set arbitrarily, without considering its effect on the accuracy of the accumulated rain estimation. In this work, {\color{blue}{we apply a novel method that sets}} a detection threshold {\color{blue}{to optimize the estimation after detection and apply it for accumulated rain estimation. }} It is based on optimizing a post-detection estimation risk function that incorporates both the estimation and the detection-related errors{\color{blue}{, which in essence takes into consideration the coupling of the detection and the estimation stages, and thus, optimizes the overall accumulated rainfall estimation}}. The proposed approach has been applied to actual CMLs attenuation measurements taken from a cellular network in Gothenburg, Sweden. It is shown that the proposed method achieves better accuracy of accumulated rain estimation, when compared with the case where the detection threshold is set independently.