Jamming recognition is a significant prior step to achieving effective jamming suppression, and the precise results of the jamming recognition will be beneficial to anti-jamming decisions. However, as the electromagnetic environment becomes more complex, the received signals may contain both suppression jamming and deception jamming, which is more challenging for existing methods focused on a single kind of jamming. In this paper, a recognition method for compound jamming based on a dual-channel neural network and feature fusion is proposed. First, feature images of compound jamming are extracted by the short-time Fourier transform and the wavelet transform. Feature images are then employed as inputs for the proposed network. During parallel processing in dual-channel, the proposed network can adaptively extract and learn task-relevant features via the attention modules. Finally, the output features in dual-channel are fused in the fusion subnetwork. Compared with existing methods, the proposed method can yield better recognition performance with less inference time. Additionally, compared with existing fusion strategies, the fusion subnetwork can further improve the recognition performance under low jamming-to-noise ratio conditions. Results with the semi-measured datasets also verify the feasibility and generalization performance of the proposed method.

A method based on data extrapolation is proposed to enhance Range-Doppler resolution of ground-based radar. The method employs both the bandwidth extrapolation (BWE), a well-known range super-resolution technique, and the timewidth extrapolation (TWE), an interesting extension of BWE from the fast-time-frequency to slow-time, to construct the radar returns with larger bandwidth and more pulses. This method consists of two key parts, firstly performing TWE on the echoes after pulse compression and then performing BWE on the integrated echoes, which can bring better performance because the gain in radar signal processing is fully utilized. Not only this, an improved data extrapolation technique is designed and applied to the proposed method. The improvement is the generalization of the existing linear prediction-based extrapolation method, and can ensure that radar resolution is further enhanced if a parameter of the suppression factor is set reasonably. The experiments based on the simulated data and the observations of vehicles on highway by an S-band ground-based radar illustrate that radar resolution can be improved by a factor of at least 3 with almost undistorted Range-Doppler images and clutter narrowed in Doppler using a conservative suppression factor setting of 0.5~1.

A non-coherent integration detection method is proposed to solve the detection and estimation of the high-speed multiple targets in radar. The method processes the targets in the real echo image within a coherent processing interval into two-dimensional peaks, based on which target detection and parameter estimation are carried out. Different from the conventional “Range-Doppler” echoes, the outputs of the proposed method are “Range-Scaled Doppler” echoes, in which the scale is set by yourself according to the maximum Doppler ambiguous number of the targets to be detected by radar, and less than 1 to ensure that Doppler frequencies of the high-speed targets are compressed to less than the pulse repetition frequency of radar. Compared with the existing keystone transform and Radon Fourier transform-based coherent methods, the proposed method does not need to search for the ambiguous number of each target, more suitable for multiple target detection. By remodeling the target's trajectory, the method achieves better performance than the existing Hough transform and Radon transform-based non-coherent methods. The experiments with the simulated and measured data illustrate the feasibility and effectiveness of the proposed method.

Echolocating bats are known to vary their waveforms at the phases of searching, approaching, and capturing the prey. It is meaningful to estimate the parameters of the calls for bat species identification and the technological improvements of the synthetic systems, such as radar and sonar. The type of bat calls is species-related, and many calls can be modeled as hyperbolic frequency- modulated (HFM) signals. To obtain the parameters of the HFM-modeled bat calls, a reversible integral transform, i.e., hyperbolic scale transform (HST), is proposed to transform a call into two-dimensional peaks in the “delay-scale” domain, based on which harmonic separation and parameter estimation are realized. Compared with the methods based on time-frequency analysis, the HST-based method does not need to extract the instantaneous frequency of the bat calls, only searching for peaks. The verification results show that the HST is suitable for analyzing the HFM-modeled bat calls containing multiple harmonics with a large energy difference, and the estimated parameters imply that the use of the waveforms from the searching phase to the capturing phase is beneficial to reduce the ranging bias, and the trends in parameters may be useful for bat species identification.