High-precision IFM receiver based on random forest algorithm

The six-port network-based Instantaneous Frequency Measurement (IFM) receiver is widely utilized in industrial applications for signal frequency demodulation. However, current IFM receivers face the issue of low algorithmic demodulation accuracy (ACC). According to the four-channel output characteristics of the IFM receiver, we use the Random Forest algorithm to extract features, train and test the output voltage data set, and recognize frequency labels. The experimental results show that the ACC of the proposed algorithm reaches 0.9583, which is 13.99% higher than that of the traditional polynomial fitting algorithm. To evaluate the noise reduction performance of the algorithm, we add Gaussian white noise to the original data and demodulate the signal with noise. Comparative tests prove the advantages of the algorithm in high stability and flexibility. Finally, multiple model evaluation indicators demonstrate that the algorithm features strong applicability and robustness to the experimental data with IFM receiver.