Identification Approach of Arriving Wave Model Based on Likelihood Ratio Test With Different Sensor Noise Levels

Abstract

Abstractk‐vector direction determined by propagation characteristics is crucial for understanding the global features of space plasma. Identifying the arriving wave model is a major factor in obtaining fast and accurate results in the direction finding of various plasma waves. If we can determine whether the observation data contain a significant natural wave or not, we can reduce the computational time for direction finding analysis by excluding noise‐only data. The conventional approach for identifying the arriving wave model assumes that all electromagnetic field sensors have same noise levels. However, the noise levels of electromagnetic field sensors on board scientific satellites can change owing to sensor degradation during long‐term instrument operation. Thus, the arriving wave model should be identified even when the noise levels of all electromagnetic field sensors are not equal. We proposed robustly identifying the arriving wave model by introducing a noise integration kernel that includes information about noise level ratios. Our proposed approach classifies a spectral matrix into three cases: noise model, single plane wave model, and multiple waves model. The proposed approach comprises the likelihood ratio test, and the identification result based on a statistical viewpoint. We conducted Monte Carlo simulations, and it was verified that the proposed approach can correctly derive the arriving wave model with high accuracy even with different sensor noise levels.