Robust multiple channel scanning and detection of low probability of intercept (LPI) communication signals

Abstract

In this paper, multiple channel scanning and detection methods for low probability of intercept (LPI) signals in a multisignal environment are proposed. The short-time Fourier transform (STFT) and short-time Goertzel algorithm (STGA) are implemented to obtain the time-frequency (TF) representation. From the TF representation, different moving average (Mavg) lengths are used to detect the presence of the signal and their performance is verified. From the receiver operating characteristics (ROC), it is shown that the detection performance of STFT and STGA are close in terms of probability of detection (Pd) using the constant false alarm (CFAR) theorem at different probability of false alarms. STFT and STGA have Pd=0.9 using Mavg=32 at signal to noise ratio (SNR) of -15 dB and -14 dB respectively. However, STGA has computational complexity (CC) that is 10 times less than STFT. The instantaneous frequency (IF) estimated from the TF representation is able to identify the signals in a multisignal environment at SNR of 5 dB.