Description: Actuality of improving of the methods for recognition of digital modulation of telecommunication systems is substantiated in the article. The analysis of known approaches to recognition of the type of modulation on the basis of statistical quantities of higher orders is carried out. It is shown that as modulation features for modulation types recognition can be used higher orders joint cumulants, which allow to synthesize algorithms for digital radio signals recognition against the background of Gaussian noise with a probability of correct recognition close to one. Analytical expressions for the calculation of joint cumulants up to the eighth order inclusive from the values of joint moments for the most common types of radio signals with digital linear modulation are obtained. The means of the statistical functions application toolbox "Statistics Toolbox" of the interactive programming environment MATLAB are extended. The procedures of joints cumulants calculation are optimized to minimize the number of mathematical operations. Recommendations for the synthesis of algorithms for modulation type recognition based on cumulant analysis are developed. The obtained results can be used to develop algorithms for signal recognition in radio monitoring systems, software-defined radio and electronic warfare.
Keywords: telecommunication system, radio signal, modulation, phase constellation, method, recognition, statistics, cumulant, moment