Detection Method of Solar Radio Bursts Based on Support Vector Machine Model

The noise signal produced by solar radio bursts (SRBs) is one of the significant factors influencing the navigation signal. The study of the detection of SRBs is meaningful to the steady operation of navigation system. Taking advantage of the influence of SRBs on Global Navigation Satellite Systems (GNSS) signal, a SRBs detection method based on support vector machine (SVM) model has been proposed. First, the carrier-to-noise ratio, positioning errors of three directions, geometric dilution of precision (GDOP) and the number of satellites loss-of-lock are input to preprocess to obtain the eigenvector and mark whether the SRBs occurs. Then the optimal classifier is obtained by inputting the sample points into SVM classifier for learning. When the new eigenvectors enter the classifier, it will be classified automatically. The average accuracy of SRBs detection reaches more than 93%. This detection method of SRBs based on SVM model can be realized all-time, all-weather and with high efficiency, high accuracy and simple process.