Deep Learning-Based Drone Classification Using Radar Cross Section Signatures at mmWave Frequencies

Rui Fu,Ahmed Abdulhakim Al-Absi,Hoon-Jae Lee,Mohammed Abdulhakim Al-Absi,Young-Sil Lee,Ki-Hwan Kim

doi:10.1109/access.2021.3115805

Rui Fu, Ahmed Abdulhakim Al-Absi + Show 4 more

Open Access

https://doi.org/10.1109/access.2021.3115805

Copy DOI

Abstract

This paper presents drone classification at millimeter-wave (mmWave) radars using the deep learning (DL) technique. The adoption mmWave technology in radar systems enables better resolution and aid in detecting smaller drones. Using radar cross-section (RCS) signature enables us to detect malicious drones and suitable action can be taken by respective authorities. Existing drone classification converts the RCS signature into images and then performs drone classification using a convolution neural network (CNN). Converting every signature into an image induces additional computation overhead; further CNN model is trained considering fixed learning rate. Thus, when using CNN-based drone classification under a highly dynamic environment exhibit poor classification accuracy. This paper present an im- proved long short-term memory (LSTM) by introducing a weight optimization model that can reduce computation overhead by not allowing the gradient to not flow through hidden states of the LSTM model. Further, present adaptive learning rate optimizing (ALRO) model for training the LSTM model. Experiment outcome shows LSTM-ALRO achieves much better drone detection accuracies when compared with the existing CNN-based drone classification model.

Highlights

Every new technology invented brings about positive and negative impacts to modern civilization
Receiver operating curve performance metrics such as accuracies, precision, recall, and F1-score are used for analyzing the drone classification model using long short-term memory (LSTM)- adaptive learning rate optimizing (ALRO)
Similar to [13] in [19] collected signature through polarimetry Ku-band frequency modulated continuous wave (FMCW) radar system. They converted the entire signal information into the image and these images were trained using a convolution neural network and the classification of the drone is performed. These models induce huge computation overhead as it collects a large amount of radar cross-section (RCS) data and that convert into an image and performs classification using GoogLeNet and CNN deep learning models

Summary

INTRODUCTION

Every new technology invented brings about positive and negative impacts to modern civilization. In [13], studied quasi-3D RCS signatures of a diverse range of drone models over the frequency band of 26-40 GHz, which may provide essential material for the drone database [18], [19] Using this measurement this work aim at developing an automatic detection drone using deep learning technique [25], [26]. For improving drone classification accuracies in this paper we present the adaptive learning rate optimization (ALRO) model. The RCS is measured considering the different frequencies and these measurement values are trained using LSTM-ALRO (Long short-term memoryAdaptive learning rate optimization) model in order to carry out drone classification tasks. For addresses in the sub-section, we introduce an adaptive learning rate optimization model for achieving better drone classification accuracies. The wwjj converges and satisfies the minimal value of optimization function; this, aid in achieving better drone detection classification accuracies

SIMULATION RESULT AND ANALYSIS

MEASUREMENT MODEL

Findings

CONCLUSION