Low-earth orbit (LEO) satellites move at very high speeds and are able to circumscribe the planet several times in a single day. Establishing communication with these systems is a multi-task process, and frequency synchronism is one of them. It’s responsible for performing a fine acquisition of the carrier frequency. This work introduces a new method of frequency acquisition for LEO satellites based on parallel searching in a set of correntropy coefficients. This system can be used in the development of a full decoder for signals from Brazilian Environmental Data Collection System (SBCDA) and ARGOS. These high dynamics environments induces considerable values of Doppler shift and Doppler rate at space communications due the variation of angle of view, nominated as a dynamic Doppler shift. The problems for frequency estimation has been discussed for years. Just a few papers consider the contributions about impulsive noise influence made by (Button et al. 2002) where he presents measurements for a more faithful characterization for LEO scenarios. On the other hand, correntropy has vastly different properties compared with second-order statistics that can be very useful in non-gaussian signal processing, especially in the impulsive noise environment. We elucidate the robustness of correntropy at presence of outliers by similarity analysis and compare with correlation. In addiction, the method gathers the energetic benefits of computing paralelism to reduces the complexity of the technique.