Comparative Analysis of Radar Cross Section Based UAV Classification Techniques

Abstract

This work investigates the problem of unmanned aerial vehicles (UAVs) identification using their radar crosssection (RCS) signature. The RCS of six commercial UAVs are measured at 15 GHz and 25 GHz in an anechoic chamber, for both vertical-vertical and horizontal-horizontal polarization. The RCS signatures are used to train 15 different classification algorithms, each belonging to one of three different categories: statistical learning (SL), machine learning (ML), and deep learning (DL). The study shows that while the classification accuracy of all the algorithms increases with the signal-tonoise ratio (SNR), the ML algorithm achieved better accuracy than the SL and DL algorithms. For example, the classification tree ML achieves an accuracy of 98.66% at 3 dB SNR using the 15 GHz VV-polarized RCS test data from the UAVs. We investigate the classification accuracy using Monte Carlo analysis with the aid of boxplots, confusion matrices, and classification plots. On average, the accuracy of the classification tree ML model performed better than the other algorithms, followed by the Peter Swerling statistical models and the discriminant analysis ML model. In general, the classification accuracy of the ML and SL algorithms outperformed the DL algorithms (Squeezenet, Googlenet, Nasnet, and Resnet 101) considered in the study. Furthermore, the computational time of each algorithm is analyzed. The study concludes that while the SL algorithms achieved good classification accuracy, the computational time was relatively long when compared to the ML and DL algorithms. Also, the study shows that the classification tree achieved the fastest average classification time of about 0.46 ms.

A Survey on Radar Techniques for Detection,Tracking, and Classification of Aerial Threats

“…The RCS of an aerial vehicle depends on the frequency, polarization, angle of illumination, and geometry and electrical properties of the material of the aerial vehicle. The effect of the frequency on the RCS measurements is provided in [26]- [28]. The RCS can be changed by changing the polarization as given in [29].…”

“…The RCS of an aerial vehicle can be used for classification. In [26], RCS of six different types of UAVs were measured at 15 GHz, and 25 GHz. The measurements were carried out in the anechoic chamber.…”

“…The Fig. 2: The measurement setup in [26] is shown. The measurements were conducted in the anechoic chamber RCS of four UAVs measured at 15 GHz and 25 GHz and at respective polarization pairs of vertical-vertical and horizontalhorizontal are provided in Fig.…”

“…The classification is obtained for low SNR detection. In [26], a comparative analysis of UAV classification techniques based on the RCS is provided. The RCS of six different types of UAVs is measured at 15 GHz and 25 GHz in an anechoic chamber.…”

“…Different classification algorithms from machine learning, statistical learning, and deep learning were used. For example, assuming equal a priori probabilities for each of the UAV type or class, the statistical rule for UAV classification given in [26] as…”