Neural network self-optimization method for modulation recognition of MFSK digital signal subclasses

Abstract

The invention discloses a neural network self-optimization method for MFSK digital signal subclass modulation recognition, including: demodulating the received modulated signal to obtain a discrete data sequence; normalizing the complex sequence; extracting the instantaneous velocity, calculating Find out the discrete map of speed changing with time; smooth the discrete map, find the smooth map of speed with time as the independent variable; solve the stable speed point and project it to the distribution density plane; train the convolutional neural network classifier model for Classify the MFSK subclass modulation signals; when the confidence is lower than the preset threshold, strengthen the smoothing effect, and execute the subsequent steps in sequence until the recognition confidence meets the requirements. The present invention can realize the modulation recognition of MFSK digital signal under actual communication conditions such as low signal-to-noise ratio and insufficient sampling. At the same time, the design of self-detection precision threshold and adjustment of mapping smoothness enables the method to have the function of self-optimization precision.

Description

technical field [0001] The invention relates to the technical field of radio communication, in particular to a neural network self-optimization method for MFSK digital signal subclass modulation identification. Background technique [0002] The identification of modulation mode of digital communication signal is an important subject of signal processing research, which is widely used in military and civilian fields. With the rapid development of communication technology, the system and modulation styles of communication signals have become more complex and diverse, and the signal environment has become increasingly dense, making it difficult for conventional identification methods and theories to adapt to actual needs and effectively identify communication signals. In recent decades, people have made a lot of beneficial explorations in the identification of communication signals, and put forward many new ideas and methods. At present, the commonly used algorithms for modulat...

AI Technical Summary

Problems solved by technology

[0008] (1) For the identification method using traditional characteristics such as signal frequency, amplitude, and phase, it does not take into account the factors of dynamic changes such as carrier frequency and amplitude;

[0009] (2) For the identification method based on the constellation diagram, it is difficult to completely filter out the slight frequency offset, and the final cumulative frequency offset effect is not considered;

[0010] (3) For the identification method based on short-time Fourier transform, the influence of factors such as noise in the actual communication environment is not considered;

[0011] (4) For the identification method using the high-order cumulant of the signal, the factor of calculation amount is not considered

[0012] At the same time, the existing methods have not considered the situation of signal loss when the ratio of code rate and sampling rate is unreasonable, and the problem of recognition accuracy under low signal-to-noise ratio

Images