Real-Time RFI Processor for Future Spaceborne Microwave Radiometers

Abstract

Anthropogenic radio frequency interference (RFI) within microwave radiometer bands is a serious problem in remote sensing. The problem is ever-increasing and thus, future spaceborne microwave radiometers will require efficient methods to mitigate the effects of RFI. In this paper, we present a solution: an RFI processor to detect and blank the RFI in real time. The processor was designed to be compatible with the 18.7-GHz channel of the Microwave Imager Instrument onboard the European MetOp Second Generation satellite system. The RFI processor developed here applies the following complementary detection algorithms: 1) anomalous amplitude detection; 2) kurtosis; and 3) two different cross-frequency algorithms. In the processing, the data are divided into subsamples with fine temporal and frequency resolution. The RFI processor detects and filters out RFI with this fine resolution in real time and then integrates the clean (noncontaminated) subsamples over time and frequency. Thus, a cleaned version of the radiometer data can be downlinked at a traditional low data rate. The processing is implemented in a reprogrammable field programmable gate array with high processing capacity, which provides high flexibility. The processing bandwidth that is applied is 200 MHz (+ 25-MHz transition bands on both sides). The detection limits of the system vary between various kinds of RFI but are in line with simulations. The power consumption of the RFI processor is ∼12 W (at room temperature), and its mass is ∼1.1 kg.