Forward Scatter Radar Moving Target Detection via Linearly Weighted Time–Frequency Entropy

Forward scatter radar (FSR) can enhance target echo signal power by exploiting the sharp increase in radar cross-section (RCS), and has been widely studied in passive radar target detection. Traditional FSR detectors operate based on the shadowing effect that occurs when a target crosses the baseline. However, when satellite transmitters are used, the probability that a target’s trajectory intersects with the baseline in three-dimensional space approaches zero. Therefore, shadowing is difficult to occur. A moving-target detection method using weighted time-frequency (TF) entropy fusion is proposed in this paper for scenarios where targets move near the baseline. First, an echo signal model is established to show that the frequency change can be approximated as linear within a short time. Then, four TF entropy features are extracted from the received signal and linearly weighted to form the test statistic. The weights are optimized using the Nelder–Mead algorithm, with the objective of maximizing the average detection probability. Finally, the effectiveness of the proposed algorithm is verified through simulations and anechoic chamber measurements. The weighted fused TF entropy achieves a higher detection probability than any single TF entropy. Compared with the energy detector, the required signal-to-noise ratio (SNR) is reduced by about 3 dB to achieve the same detection probability at a false alarm probability of 10−3.