LPI Detection of Frequency-Hopping Signals Using Autocorrelation Techniques

Interception of frequency-hopping (FH) waveforms is commonly achieved by using a set of radiometers (energy detectors) that individually energy-detect subbands of the total spread bandwidth of the suspected transmission. In this paper, a scheme is analyzed which can improve the detection performance of FH waveforms in wide-band additive Gaussian noise (AWGN) using samples from the autocorrelation domain. It is shown that, under fairly general operational assumptions, an appreciable gain in decision signal-to-noise ratio (SNR) can be achieved over that of the radiometer. This gain is proportional to
\gamma_{H}^{2}
where
\gamma _{H}
is the hop SNR. The proposed algorithm, albeit inferior to the optimal likelihood-ratio test, has the advantage of reduced complexity. The overall approach has been motivated by the recent implementational feasibility of large time-bandwidth-product real-time correlators such as surface-acoustic wave devices (SAWD's).