Tutorials

SAR/ISAR images have been largely used for earth observation, surveillance, classification and recognition of targets of interest. The effectiveness of such systems may be limited by a number of factors, such as poor resolution, shadowing effects, interference, etc. Moreover, both SAR and ISAR images are to be considered as two-dimensional maps of the real three-dimensional object. Therefore, a single sensor may produce only a two-dimensional image where its image projection plane (IPP) is defined by the system-target geometry. Such a mapping typically creates a problem for the image interpretation, as the target image is only a projection of it onto a plane. In addition to this, monostatic SAR/ISAR imaging systems are typically quite vulnerable to intentional jammers as the sensor can be easily detected and located by an electronic counter-measure (ECM) system. Bistatic SAR/ISAR systems can overcome such a problem as the receiver can act covertly due to the fact that it is not easily detectable by an ECM system, whereas multistatic SAR/ISAR may push forward the system limits both in terms of resolution and image interpretation and add to the system resilience.

This tutorial introduces electromagnetic warfare (EW) concepts and principles necessary for modern combat systems. The focus will be on electronic support (ES) and electronic attack (EA) functions. The intent is to familiarize the audience with EW concepts and achieve an understanding of how EW is used to interrupt radar processing chains. This talk covers a general discussion on the EW field, including applications outside radar-specific uses and terminology widely used within the field. A historical development of the EW field will be presented to motivate importance and historical use. Basic EW techniques (e.g. noise, range/velocity techniques, etc.) with associated effects on nominal radars will be presented/discussed to ensure an understanding of the technical underpinnings of EW. Building on the basic techniques, a brief discussion on concepts in advanced EW systems and current research will be presented. The discussion will conclude by briefly discussing the revolutionary impact of cognitive and AI/ML processes on EW.

Modern radar systems with wide-bandwidth waveforms tracking targets with increasingly complex dynamics require state-of-the art track filters to make best use of the full potential of the radar system accuracy. In many cases, the standard extended Kalman Filter (EKF) algorithm is insufficient to provide good estimation performance due to the non-linearity of the dynamics and/or measurement functions or the existence of multiple dynamics modes. While particle filters provide a catch-all approach to these difficulties, they come with great computational expense due to the so-called "curse of dimensionality," a phenomenon in which the number of particles or parameters associated with a given state representation must increase exponentially with the state dimension in order to achieve good performance. For these complex filtering problems, a solution that provides both good estimation performance and reasonable runtime is sought. Interacting multiple model (IMM) filters, sigma-point filters and Gaussian mixture filters are approaches that have shown promise regarding these goals. This tutorial will present the background needed to understand and apply both these algorithm types to complex estimation problems relevant to radar systems. Numerical examples are used to illustrate the application. Novel techniques for improving the runtime efficiency of sigma-point filters are also covered. The techniques include a discussion of selecting the correct order sigma-point rule to minimize runtime for a given estimation problem. Regarding Gaussian mixture filters, the tutorial focuses on the contact-lens problem in radar. The limitations of sigma-point filters to address the contact-lens problem and recent improvements achieved by applying a Gaussian mixture estimation approach are discussed.

A modern multi-function phased array radar system can handle several tasks by rapidly pointing beams into different directions, adapting waveform, frequency and transmitted power, splitting its array antenna into sub-apertures or coordinating with other radars. Human operators will struggle to manage the available options coming with enhanced radar capabilities, necessitating automated management techniques. These techniques, driven by recent advances in optimization and robotics, are becoming crucial for the performance of the next generation of multi-function radar systems. This tutorial provides an introduction to radar resource management, describing classical as well as recent solutions to the challenge presented by managing ever more capable radar systems. It will teach a framework to encode the utility of multiple tasks into an objective that can be understood by a radar manager and explain key concepts like adaptive tracking, Quality of Service and its optimality conditions. Additionally, the tutorial discusses how aspects like cognitive radar, multi-sensor management and machine learning impact radar resource management.

The introduction to the course provides a history of AESA development going back to the 1960s. Differences between mechanically scanned arrays (MSAs), passive electronic scanned arrays (PESAs) and AESAs are discussed and elaborated upon in addition to the progressive advancement of AESAs. Also included is a summary of the benefits that AESAs provide for different applications. These applications include Radar, Electronic Warfare (EA/ESM), SIGINT and COMMS. Finally, using the radar range equation, the impacts of array elements, transmit receive modules (TRMs), and beamformers on system performance is detailed.

This tutorial focuses on passive radar and illustrates the many successful methods and amazing solutions that can be adopted in such sensors to increase their reliability, improve their potentialities, and hence widen the range of applications. The tutorial starts from the basic concepts of passive radar by discussing the possible illuminators of opportunity, the impact of the bi/multi-static geometry, as well as the passive radar equation. A typical signal processing scheme is introduced, and effective solutions are illustrated for the signal processing techniques to be implemented at each stage, there including clutter filtering, cross-ambiguity function calculation, target detection, direction of arrival (DoA) estimation, bistatic/Cartesian tracking. Ground-based passive radar systems are first investigated, including the demonstrator and operational system design. Mission planning, which involves the calculation of the passive radar performance (detection range and accuracy), is also covered. Therefore, advanced methods are illustrated to enhance the performance of the passive radar sensor by exploiting long integration times, the polarization/frequency/spatial diversity provided by using multiple channels on receive, or the target scattering observed under moderate to extreme bistatic geometries. Then, the discussion focuses on advanced operative modes for passive radar. Specifically, the possibility of installing a passive radar onboard moving platforms is considered, which enables the capability of forming images of the surveyed scene as well as the possibility of a stand-off surveillance of ground moving targets. Therefore, the principle of operation, the signal models, and the signal processing techniques are illustrated with reference to both ground surface imaging in SAR mode and passive radar GMTI. Moreover, the last tutorial section is dedicated to target imaging with passive radar in ISAR mode and includes a description of the required methods as well as the challenges to be faced. In addition to the theoretical aspects, the tutorial provides the attendees with an insight into the real-world applications of passive radar. A wide range of applications is covered, such as air traffic control, including surveillance against UAVs, maritime surveillance, vehicular traffic monitoring, to indoor surveillance and, for each case, several experimental results are reported exploiting different illuminators of opportunity (FM radio, DVB-T, WiFi, etc.). Walking through these results gives the chance to describe in more detail some technical aspects related to system design issues and signal processing techniques, as well as to understand the current limitations and future perspectives of passive radar sensing.

This tutorial will provide a first-principles examination of the design goals and metrics of both radar and communications. We will explore the motivation and history of spectrum access and examine the practical requirements for utilizing the available DoFs. Specific examples of coexistence and co-design techniques will be explored based on the DoF(s) they use to enable efficient spectrum access. For example, the problem space of coexistence of radar and commercial communications will be explored in detail – from problem setup, to system requirements, to demonstrations of real-time processing. For the co-design problem two distinct families of techniques will be framed and explored in detail: radar-embedded communications via coding diversity and multi-beam emissions from digital arrays. Implications of hardware constraints on these techniques will be illustrated. To narrow the focus, radar detection will be the primary radar application.

Traditional maritime radar is based on non-coherent detection, mainly due to the complexities of implementing coherent detectors in sea clutter. Over the past decade, there has been significant new research into the characterisation and modelling of sea clutter and how to improve maritime target detection. The use of models has also led to techniques for predicting the performance of many new radar detection schemes. This tutorial will include a comprehensive coverage of new research in three key areas. The first is sea clutter modelling and its application to target detection. The second area looks at several detection schemes that have been proposed for detection of targets in sea clutter. These include both non-coherent techniques based on constant false alarm rate (CFAR) schemes, coherent single and multichannel techniques. The final part of the tutorial looks at several new techniques for target detection, including approaches based on time-frequency analysis, sparse signal separation, machine learning and track-before-detect.

Adaptive beamforming is driving the adoption of larger, all-digital electronically steerable arrays (ESAs/phased arrays). However, the underlying concepts and mathematics can often feel abstract and challenging to grasp. In this hands-on workshop, we will bridge that gap by building and implementing our own multi-channel adaptive beamformers. Participants will witness it in action and adaptively manage jammers and interferers. We will methodically cover the fundamentals of adaptive beamforming, exploring various implementations step-by-step. Participants will then design their own algorithms to analyze real data collected from a digital beamformer available in the room. Each topic will include a concise lecture explaining the relevant theory and mathematics, followed by practical, hands-on activities using real-world data. 
To facilitate participation, all necessary data and Python scripts will be provided during the tutorial, allowing attendees to perform the labs directly on their laptops. This interactive approach ensures participants gain a deeper, intuitive understanding of adaptive beamforming through both theory and practice.

The focus of this tutorial will be hands on implementation (laboratory) and theory of machine/deep learning for radio frequency automatic target recognition (ATR). For this tutorial, the author will use his recently published (July 2020) book by Artech House "Deep Learning for Radar and Communications Automatic Target Recognition". This authoritative resource presents a comprehensive illustration of modern Artificial Intelligence / Machine Learning (AI/ML) technology for radio frequency (RF) data exploitation.

The purpose of this tutorial is to provide a serious exposition of the state-of-the-art of passive radar and its development in the context of target detection and imaging. This tutorial will be pitched so as to present bistatic and multistatic passive radar using novel wideband illuminators of opportunity in an advanced format. The tutorial will focus on developing the grounding of advanced principles and concepts that are, and will be, of high relevance to the field. This tutorial will be of high value to scientists and engineers working with passive radar technology, representatives of the military, government and industry and to other postgraduates involved in the field of radar as well as seasoned practitioners. Our goal is the delivery of modern advanced topics in an accessible format. By the conclusion of the tutorial, participants will have acquired a deep appreciation of core advanced topics relating to passive radar using new wideband illuminators of opportunity, such as 5G/6G, WiFi, DVB-S and Fixed Satellite Services (FSS, such as STARLINK and OneWeb), and the required signal processing techniques. The tutorial will include different standards comparison, challenges, opportunities and limitations analyzes with focus on modern applications for using wideband IoOs in passive radars, e.g., target detection, classification, SAR/ISAR imaging that participants would not have accrued through self-study of recently published literature. Representative examples will be used throughout the tutorial to aid understanding. Worked examples with interactive participation will ensure a lively tutorial for the full duration.

We teach advanced radar detection from first principles and develop the concepts behind Space-Time Adaptive Processing (STAP) and advanced, yet practical, adaptive algorithms for realistic data environments. Detection theory is reviewed to provide the student with both the understanding of how STAP is derived, as well as to gain an appreciation for how the assumptions can be modified based on different signal and clutter models. Radar received data components are explained in detail and the mathematical models are derived so that the student can program their own MATLAB or other simulation code to represent target, jammer and clutter from a statistical framework and construct optimal and suboptimal radar detector structures. The course covers state-of-the-art STAP techniques that address many of the limitations of traditional STAP solutions, offering insight into future research trends.

Inverse Synthetic Aperture Radar (ISAR) is a well-known technique to obtain high-resolution radar images of non-cooperative targets. ISAR images have been largely used to classify and recognise targets and ISAR technology is nowadays employed and integrated in modern radar systems. Nevertheless, despite decades of research and development work in ISAR imaging, two-dimensional (2D) ISAR images present some intrinsic drawbacks that limit the effectiveness of their use for target classification and recognition. Some of these limitations come from the unpredictability and uncontrollability of the image projection, which transforms three-dimensional (3D) targets in 2D images. One very effective way of overcoming this problem is to form 3D ISAR images instead of 2D ones. 
This tutorial will present a unique walkthrough 3D ISAR imaging, including concepts, algorithms, systems and real data examples, which will provide the attendants the necessary tools for a full understanding of this new technology.

Autonomous driving is one of the megatrends in the automotive industry, and a majority of car manufacturers are already introducing various levels of autonomy into commercially available vehicles. The main task of the sensing suite in autonomous vehicles is to provide the most reliable and dense information on the vehicular surroundings. Specifically, it is necessary to acquire information on drivable areas on the road and to port all objects above the road level as obstacles to be avoided. Thus, the sensors need to detect, localize, and classify a variety of typical objects, such as vehicles, pedestrians, poles, and guardrails. Comprehensive and accurate information on vehicle surroundings cannot be achieved by any single practical sensor. Therefore, all autonomous vehicles are typically equipped with multiple sensors of multiple modalities: radars, cameras, and lidars. Lidars are expensive and cameras are sensitive to illumination and weather conditions, have to be mounted behind an optically transparent surface, and do not provide direct range and velocity measurements. Radars are robust to adverse weather conditions, are insensitive to lighting variations, provide long and accurate range measurements, and can be packaged behind the optically nontransparent fascia. The uniqueness of automotive radar scenarios mandates the formulation and derivation of new signal-processing approaches beyond classical military radar concepts. The reformulation of vehicular radar tasks, along with new performance requirements, provides an opportunity to develop innovative signal processing methods. This Tutorial will first describe active safety and autonomous driving features and associated sensing challenges. Next, it will overview technology trends state advantages of available sensing modalities and describe automotive radar performance requirements. It will discuss propagation phenomena experienced by typical automotive radar and radar concepts that can address them. It will compare radar and LiDAR signal processing chains and emphasize their similarity, differences, and associated processing challenges. Next, this Tutorial will focus on the radar processing chain: range, Doppler measurement estimation, beamforming, detection, range and angle-of-arrival migration, tracking, and clustering. Discussing modern automotive radars, the Tutorial will describe the MIMO radar approach. Finally, the automotive radar applications and advanced topics, such as interference mitigation and sensor fusion, will be discussed.