Deep Learning for Radar and Communications Automatic Target Recognition presents a comprehensive illustration of modern Artificial Intelligence / Machine Learning (AI/ML) technology for radio frequency (RF) data exploitation. While numerous textbooks focus on AI/ML technology for non-RF data such as video images, audio speech, and spoken text, there is no such book for RF data. Hence, there is a need for RF machine learning textbook for the research community that captures state-of-the-art AI/ML and deep learning (DL) algorithms and future challenges. The book provides the practitioner with: (i) an overview of the important ML/DL techniques, (ii) an exposition of the technical challenges associated with incorporating ML methods to RF applications, and (iii) implementation of DL techniques on synthetic aperture radar (SAR) imagery and communication signals classification.
The book highlights a comparative analysis of various machine learning and deep learning techniques applicable to radar and communications. Simplified examples and diagrams give the reader further insight into the similarities and differences between the many ML strategies that can be developed and utilized. Furthermore, the book gives a comprehensive reference list for each of the approaches within each chapter, so that the reader is easily steered towards additional auxiliary information to accomplish needed goals on radar and communications image and signal classification problems. Read on for an in-depth look at some of the problems this book helps solve:
Radar/Radio Frequency Image classification for finding targets/objects:
Many surveillance radars are built on synthetic aperture radar (SAR) which generate images of a ground scene. From the images, the challenge is to detect targets and activities. For example, people may drive cars to an unauthorized place/facility. After collecting the images, modern Artificial Intelligence/Machine Learning/Deep Learning algorithms classify the vehicles, human, and the surroundings. The textbook “Deep Learning for Radar and Communications Automatic Target Recognition” illustrates how deep learning can be applied to classify various targets (military and civilian vehicles, surroundings) from SAR imagery. The book presents many historical machine learning algorithms (from early days to modern deep learning algorithms) applied SAR imagery and classification performance analysis of these methods. The book helps technical professionals to understand (implementations, pros/cons, technical issues) many DL algorithms in a single book.
Digital Communication Signal Classification:
Digital communication signals include modern IoT (Internet of Things) devices such as Wi-Fi and Bluetooth Devices, mobile phones, and personal hotspots. Other signals of interest include aircraft communication signals, radar signals, TV signals. Satellite communication signals. Radio Frequency (RF) fingerprinting involve identifying RF devices based on their unique hardware signatures. The book “Deep Learning for Radar and Communications Automatic Target Recognition” discusses how modern machine learning can be used to classify various communication signals. In the past, these classifications have been implemented using hand-crafted, inefficient algorithms that don’t scale to large numbers of devices identifications. As previous machine learning algorithms find it difficult to identify new devices that have been recently included within the measured data, deep learning approaches help to identify communication signatures in real time.
Performance analysis of Automatic Target Recognition (ATR) algorithms
This book provides theory and implementation of ATR performance analysis. Often researchers develop Artificial Intelligence / Machine Learning based target classification algorithms that perform very well on some datasets but are not robust to all situations. ATR performance analysis is discussed so as to understand limitations and benefits of machine learning algorithms in various operating environments. The book describes ATR performance analysis for designers, developers, and practitioners.
Click here for more information or to order.