Identification of Shortwave Radio Communication Behavior Based on Autocorrelation Spectrogram Features
Table 2
The process of extracting autocorrelation spectrogram features.
Step
Detailed process
Step 1
signal generation: According to the MIL-STD-188-141B standard, generate five types of burst waveforms with noise
Step 2
calculate the autocorrelation function: Based on Equation (5), calculate the autocorrelation function , to achieve the purpose of noise reduction
Step 3
bispectrum transformation: According to Equation (6) and (7), calculate the bispectrum transformation matrix of . The dimension of is
Step 4
down sampling: For matrix , sample one sample point for every two points, so that the matrix dimension becomes 256128, to realize the dimension reduction of the matrix
Step 5
constructing a three-dimensional matrix: Extract the odd and even rows of the matrix, respectively, and construct matrices and with a dimension of . and are the first and second channels of the three-dimensional matrix , and the dimension of becomes