Christos Ilioudis

This article presents a comprehensive review of micro-Doppler processing techniques in synthetic aperture radar (SAR). The analysis of the micro-Doppler effect in SAR is becoming an important area of research that has the potential to unlock a new layer of useful information available within SAR images. The article begins by laying a strong theoretical foundation before presenting and discussing the challenges associated with accurately extracting micro-Doppler information from SAR data: reviewing the signal processing techniques and collating some of the main results achieved to date. The article concludes with an overview of the gaps and limitations in current research, underscoring areas for future investigation.

Satellite-borne Synthetic Aperture Radar (SAR) technology has revolutionized remote sensing applications by providing high-resolution and all-weather imaging capabilities. With the increasing availability of SAR data, the need for efficient data processing has become crucial. On-board processing has emerged as a promising solution to address the challenges associated with limited downlink capacity and final data products latency. Performing data processing and compression directly on the airborne platform reduces the raw data transmitted to ground stations, which offers several key benefits. Firstly, it significantly reduces the data volume to be downlinked, optimizing the usage of limited bandwidth and minimizing transmission delays. Secondly, on-board processing enables faster access to processed data products, allowing for quicker decision-making and timely response to dynamic events, enhancing the real-time capabilities of the system that are particularly valuable in time-critical applications. This article discusses various on-board processing techniques employed in SAR systems and explores their challenges, like computational constraints, architectural impacts, power consumption, and algorithm optimization. Furthermore, it examines the potential future developments in on-board processing, such as the integration of artificial intelligence and machine learning techniques to enhance data analysis and decision-making capabilities. The advancements in on-board processing have the potential to revolutionize the way SAR missions are conducted. By leveraging these techniques, SAR systems can achieve improved operational efficiency, reduced data latency, and enhanced real-time capabilities. This article emphasizes the significance of on-board processing in meeting the growing demands of SAR applications and underscores its role in advancing remote sensing capabilities for various sectors, including environmental monitoring, disaster response, and surveillance.

Traditional radar sensors used for surveillance rely on monostatic radar principles. However, recently the use of remote radio frequency telescopes as bistatic receivers represents an interesting way to reuse existing facilities while providing additional information to improve tracking accuracy. In this paper we study the benefits of using such a system for the task of manoeuvre detection in satellites in LEO and MEO. We investigate the conditions in which a multistatic radar is advantageous for this purpose, and show concrete results based on simulated data. Moreover, we propose novel manoeuvre detection methods, and compare their accuracy to methods found in the literature. A more general way of assessing the accuracy of these manoeuvre detection methods is also proposed, with the aim of taking into account that the parameters of the manoeuvre that actually takes place also have an effect on the accuracy. These can be split into optimal control based methods, and statistical methods. We found the addition of multistatic radar to allow considerable improvement in the accuracy of the manoeuvre detection process, an improvement that is shown to be greater the greater the baseline, i.e., the distance of the receiver to the transmitter. Furthermore, the manoeuvre detection methods that accurately model the uncertainty in the measurements were found to be the most accurate.

Classical time-frequency distributions, as the Short Time Fourier Transform (STFT) or the Continuous Wavelet Transform (CWT), aim to enhance either the resolution in time or frequency, or attempt to strike a balance between the two. In this paper, we demonstrate how a super resolution technique, the Superlet based time frequency distribution, named Superlet Transform (SLT), can boost the performance of existing classification algorithms relying on information extraction from the micro-Doppler signature. SLT is applied to provide a time-frequency distribution with finer resolutions that would boost the performance of micro-Doppler classification approaches based on time-frequency distributions (TFDs). This work shows the effectiveness of the integration of SLT in the processing pipeline with verification on real radar data.

In recent years, the technological advances leading to the production of high-resolution Synthetic Aperture Radar (SAR) images has enabled more and more effective target recognition capabilities. However, high spatial resolution is not always achievable, and, for some particular sensing modes, such as Foliage Penetrating Radars, low resolution imaging is often the only option. In this paper, the problem of automatic target recognition in Low Resolution Foliage Penetrating (FOPEN) SAR is addressed through the use of Convolutional Neural Networks (CNNs) able to extract both low and high level features of the imaged targets. Additionally, to address the issue of limited dataset size, Generative Adversarial Networks are used to enlarge the training set. Finally, a Receiver Operating Characteristic (ROC)-based post-classification decision approach is used to reduce classification errors and measure the capability of the classifier to provide a reliable output. The effectiveness of the proposed framework is demonstrated through the use of real SAR FOPEN data.

The increasing amount of space debris poses a growing challenge in the development of space travel. In previous work, a proposed system comprising of a low budget space‐borne passive radar based on CubeSAT, flying in Low Earth Orbit (LEO), demonstrated promising results in terms of space debris detection performance. In this work, a novel shadow inverse synthetic aperture radar (SISAR) model for space debris imaging based on nanosatellite is presented. An analysis of the received forward scattering signal is provided along with a novel model for moving transmitters and receivers SISAR imaging. Using simulated data and assuming different scenarios, the performance of the new algorithm is assessed; wherein, a classification scheme is applied using the output of the SISAR algorithm for classification between simple target shapes.

As unmanned aerial vehicles (UAVs) are becoming more accessible and easier to use, the need for a reliable and easy to deploy monitoring solution has become the subject of great importance in defence, security, and commercial sectors. In this work, a novel passive bistatic radar is proposed to facilitate UAV detection and localisation by exploiting global navigation satellite systems (GNSS) as illuminators of opportunity. The result of a feasibility study conducted to determine the maximum operational range of the system under different configurations as well as its parameter estimation capabilities was evaluated through simulations. To facilitate multiple satellite signals, a multiple‐input single‐output approach is adapted to estimate the target's location and velocity.

In this work a generalised signal model is presented to accommodate both narrowband and wideband signals in a multi-input multi-output (MIMO) sensor system scenarios. The derived model is then used to define a MIMO ambiguity function (AF) based on the Kullback-Leibler divergence (KLD). Moreover, the proposed formulation is parametrised using the signal and channel correlation matrices to account for different waveform and sensor placement designs, thereby allowing a flexible modelling approach. A comparison between the proposed definition and the more conventional approach of summing the squared matched filter outputs is presented for different sensors and waveforms configurations.

Nowadays the identification of ballistic missile warheads in a cloud of decoys and debris is essential for defence systems in order to optimize the use of ammunition resources, avoiding to run out of all the available interceptors in vain. This paper introduces a novel solution for the classification of ballistic targets based on the computation of the inverse Radon transform of the target signatures, represented by a high resolution range profile frame acquired within an entire period of the main rotation of the target. Namely, the precession for warheads and the tumbling for decoys are taken into account. Following, the pseudo-Zernike moments of the resulting transformation are evaluated as the final feature vector for the classifier. The extracted features guarantee robustness against target's dimensions and rotation velocity, and the initial phase of the target's motion. The classification results on simulated data are shown for different polarizations of the electromagnetic radar waveform and for various operational conditions, confirming the the validity of the algorithm