The principles of flicker noise spectroscopy (FNS) are presented - a general phenomenological approach to the analysis of signals of different nature (time series, surface roughness profiles, chaotic spectra, etc.) and the extraction of information contained in the combination of "resonant" and "chaotic » components of the studied signals. The essence of the FNS-approach based on the concept of the irreversibility of the spatio-temporal evolutionary dynamics of open complex systems at all levels of the hierarchy is described in detail. The possibilities of the method are demonstrated by examples of applications to the problems of physics, radiophysics, astrophysics, geophysics, physical chemistry, and medicine. For physicists, chemists and biologists who are interested in the problems of the dynamics of complex systems and the analysis of the information essence of chaotic signals. The publication was supported by the Russian Foundation for Basic Research under project 07-02-07027