In the process of the target hitting the bottom, strong instantaneous signals are generated. Monitoring, processing and analyzing these signals can help to determine whether the signal is from abnormal activities or not, and analyzes relevant infor- mation according to signal to realize early warning and quick response. Since bispectrum detection has a good capability in depress- ing Gaussian interference, stationary Gaussian random processes and bispectrum characteristics of simulating instantaneous seismic wave signals produced by bottoming experiments are analyzed in this paper. Instantaneous signals acquired by test are detected and simulated in the Gaussian background, and results show that bispectrum detection can effectively detect the instantaneous signals un- der background with Gaussian interference[1] .