Purpose. The existing system of automatic locomotive signaling (ALS) was developed at the end of the last century. This system uses the principle of a numerical code which is implemented on the basis of relay engineering, and therefore, it is exposed to various types of interferences. Over the years, the system has been upgraded several times, but the causes of faults and failures in its operation are still the subject of research. It is known that the frequency and the phase modulation of signal has a higher interference immunity as compared to the amplitude modulation. Therefore, the purpose of the article is to study the possibility of using the frequency methods such as Fourier series expansion and wavelet decomposition to extract the informational component of the received code from ALS signals under the action of various types of interferences. Methodology. One can extract the information unavailable in time representation of the signal using the signal studies in the frequency domain. The wavelet decomposition has been used for this purpose. This makes it possible to represent the local characteristics of the signal and to provide time-frequency decomposition in two spaces at the same time. Due to the high accuracy of the signal representation it is possible to analyze the time localization of spectral components and eliminate interference components even in the case of coincidence of interference frequency with the signal carrier frequency. Findings. To compare informativity of the methods of Fourier expansion and wavelet decomposition it was studied the reference and noisy signal of green fire code using the software package MATLAB. Detailed analysis of the obtained spectral characteristics showed that the wavelet decomposition provides a more correct decoding of the signal. Originality. Replacing the electromagnetic relays in the ALS system by microprocessor hardware involves the use of some mathematical tool for decoding, in order to obtain more information about the code. More often than not, as a mathematical tool, the classical Fourier decomposition is used. But because of a number of drawbacks in this method, it was suggested to use the wavelet decomposition, which has a number of advantages and accounts the disadvantages of the Fourier transform. Practical value. The presented method of code signal research can be the basis for developing dynamic model of the ALS receiver and decoder using digital processing module, which will enable to increase the reliability and accuracy of extraction of the code information component.