Compressed sensing based detection of localized heavy rain using microwave network attenuation

This paper proposes a localized heavy rain detection method based on the compressed sensing algorithm and using rainfall attenuation of microwave mesh networks. The proposed method is capable of reconstructing the specific attenuation of an area from the measured link attenuations. The network system configuration for this paper used three links crossing a single area to improve the accuracy in detecting the location and intensity of rain fall. The false positive and false negative results show that the proposed method has a high accuracy in detecting both the location and intensity of rain. Furthermore, not only the localized heavy rains were detected, but the results also showed high accuracy for other rain rate categories.