Distinction of Scrambled Linear Block Codes Based on Extraction of Correlation Features

Aiming to solve the problem of distinction scrambled linear block codes, a method for identifying scrambling types codes by combining correlation features and convolution long short-term memory neural networks is proposed in this paper. First, cross-correlation characteristics sequence symbols are deduced, partial autocorrelation function constructed, superiority determined derivation, two combined as input characteristics. A shallow network convolutional LSTM constructed; finally, code dataset into model, training recognition test completed. The simulation results show that, compared with traditional algorithm based on multi-fractal spectrum, can identify synchronous scrambler, accuracy higher under high bit error rate. Moreover, suitable classification noise. lays foundation future improvements scrambler parameter identification.