1993 年 20 巻 4 号 p. 228-233
In the case of the passive sonar, the LOFAR-Grams of received signals using frequency analysis and variable density visualization technique is widely used for target signal detection. The auto-spectrum analysis method is a signal processing means for the LOFAR-Grams.
As the means to improve the ratio of signal to noise levels (S/N), there is a process of averaging (cumulative processing) analyzed signals used in the LOFAR-Grams.
The discrimination signals smoothed cumulatively in the auto-spectrum method are apt to be difficult due to the nature of noise components, and this auto-spectrum method is not used in this case.
We proposed a new method here, and analyzed experimentally several types of signals of the passive sonar utilizing this method for the LOFAR-Grams.
The results showed that the process gain by cumulative method was obtained in accordance with calculation, and the delay type cross-spectrum method gave processing gain 3 dB higher than the auto-spectrum method.
It was considered that this technique was to be applied extensively to the passive sonar in the future.