This paper explores spectral enhancement techniques for Doppler Tomography (DT) and Doppler Back Projection (DBP) in order to achieve improved 2D imaging of rotating space objects observed using bistatic radar. DT and DBP are utilized in narrowband radar systems for imaging targets rotating about single and multiple axes, respectively. However, the limited range information, the small angular displacement for coherent computations and additional rotations in the targets lead to less detailed 2D images. To address this limitation, we investigate spectral enhancement methods to obtain sharper images, enabling better Space Domain Awareness (SDA). The paper provides an introduction to key concepts such as bistatic geometry, signal modeling for rotating targets, and the processing sequence. This sequence encompasses preprocessing steps, the implementation of DT/DBP techniques, and a thorough consideration of resolution-related factors. Each technique is closely associated with specific image enhancement approaches, accompanied by a detailed analysis of their advantages and disadvantages. The effectiveness of these enhancements is quantitatively evaluated using performance metrics, and practical validation is demonstrated through real-world measurements presented in the results section.

Abstract Long baseline bistatic radar systems herald enhanced sensitivity and metric accuracy for space objects in geosynchronous orbits and beyond. Radio telescopes are ideal participants in such a system; in particular, they often feature large apertures with low‐noise temperatures and have stable, synchronised clocks. Pairing radio telescopes with high‐power radars creates new methodologies for Space Domain Awareness. This paper describes long baseline bistatic measurements using the Millstone Hill Radar in the USA, the Tracking and Imaging Radar in Germany, multiple receivers of the enhanced multi‐element remotely linked interferometer network array in the United Kingdom, and the Westerbork Synthesis Radio Telescope in the Netherlands. The authors, a Research Task Group formed by the NATO Science and Technology Organisation Sensors and Electronic Technology Panel (SET‐293), performed novel bistatic and monostatic radar imaging experiments with real on‐orbit tumbling rocket bodies. These experiments on tumbling objects at near‐geosynchronous orbits highlight successful demonstrations of advanced bistatic Doppler characterisation across diverse imaging geometries. Specialised Doppler processing on tumbling targets, such as the Doppler superpulse algorithm, enables high‐fidelity rotation period estimation and determination of minimum target size.

Bistatic radars with long baselines are gaining popularity due to the increased interest in Space Domain Awareness, but require significant investment due to the need for large antenna apertures, low noise temperatures, and synchronized clocks. However, radio telescopes can be paired with high-power radars to create long baseline bistatic-multistatic radar networks. This paper describes novel long baseline bistatic measurements using the Millstone Hill Radar (MHR) in the USA and receivers in multiple antennas of the e-MERLIN array in the United Kingdom and the Westerbork Synthesis Radio Telescope (WSRT) in the Netherlands. When provided with basic information about the transmitted waveform, multiple receivers can detect, identify, and track multiple targets. The paper also presents a framework for imaging tumbling targets, demonstrating its effectiveness on real data collected by a novel system concept. This framework integrates coherent integration and Doppler processing techniques to improve signal-to-noise ratio, estimate and correct target motion, and estimate the rotation period and dimensions of tumbling targets, resulting in a clear image of the space object. The presented framework and the proposed system concept have potential to improve bistatic imaging, leading to better characterization of objects in Earth orbits undergoing complex motion, with significant implications for applications such as space domain awareness, space traffic management, and space object characterization.

Classification modes aimed for slow events, such as human gait, using micro-Doppler measurements are long, and therefore not suitable for use during time critical radar operation. When these modes are interruptible by other radar modes they might become more acceptable. In this paper, sparse signal processing, in particular the SL0-algorithm, is investigated for interpolating interrupted radar measurements which subsequently can be used for classification of human gaits. The performance of a k-NN classifier with PCA-based feature extraction was improved significantly when the data were interpolated as compared to using the interrupted data without interpolation.

In urban and littoral environments, a wide variety of moving objects such as vehicles, birds, ultralights, drones, and small aircraft may lead to a large number of closely spaced radar detections that can cause severe problems in the tracking process such as false tracks, merged tracks and track loss. By extracting non-kinematic features such as micro-Doppler signatures or 1-D or 2-D spatial target images from the radar signal, the radar detections can be classified in different target classes to mitigate the problems in the tracking process and improve situation awareness. However, micro-Doppler signature analysis and high resolution imaging techniques need a long time-on-target that may be incompatible with the large number of targets in urban or littoral environments. Furthermore, micro-Doppler features are not always observable and target images can be distorted by moving parts and multipath on the target body. In this paper, range-Doppler imaging for air target classification is assessed.

In this section we present a feature-based approach to estimate human motion parameters from radar spectrograms. The walking model of Boulic is used with personification information of the torso and leg. We have proposed a sinusoidal model for the torso and leg. The torso sinusoidal is related to the centre velocity and the leg sinusoidal is related to the lower and upper velocity bounds. Three methods are described which extract these velocities. Sinusoidal fits of velocity-slices give the cycle frequency that originates from periodic spectrogram components. Kalman filters smooth the features. The animated human generated with the features provide a realistic look-alike of the real motion of human irrespective of the walking model used to generate it. The methods are applied to real radar measurements with different scenarios. The three methods give approximately equivalent results. The percentile method has best match with the velocity bounds and centre velocity. The computation time of the correlation method is about six times the other methods. We advise the percentile method with leg frequency and cycle frequency fusion. If enough computation time is available, the cycle frequency is not needed in real-time applications.

In this paper, a new signal processing technique for Multiple-Input Multiple-Output (MIMO) image processing of a Linear Frequency-Modulated Continuous Wave (LFMCW) automotive radar was developed. The image is the range-speed-direction data cube. The technique comprises two improvements to the standard MIMO image processing: 1) A fast unambiguous Doppler target detection. 2) An improved direction finding with estimated virtual antenna elements besides the synthesized virtual antenna elements. High speed automotive targets go through several range cells in the observation period making it difficult to obtain coherent accumulated power. Another problem with high speed targets in combination with a low Sweep Repetition Frequency (SRF) is ambiguous Doppler speed measurement. In order to remove range migration and obtain coherent accumulated power the keystone transform is applied. The direction finding resolution depends on the number of transmitter-and the receiver elements of the MIMO configuration. A combination of the transmitter and receiver elements give synthesized virtual elements with corresponding antenna width. In order to improve the direction finding, the antenna width with extrapolated virtual antenna elements was extended. The Band Width Extrapolation (BWE) method is applied for estimating these extrapolated virtual antenna elements. The proposed method is verified by simulation and real radar measurements.

A light weight SAR, suitable for use on short range tactical UAV, has been designed and built. The system consists of a fully digital receive array, and a very compact active transmit antenna. The approximate weight of the complete system is 6 kg, with power consumption below 75 W, depending on the required data acquisition volume. This X-band system is designed to provide an image resolution down to 10–15 cm at a maximum range of approximately 5 km. The design is described, and initial ground test results are presented.

This paper presents methods for Coherent Multistatic Radar Imaging for Non Cooperative Target Recognition (NCTR) with a network of radar sensors. Coherent Multistatic Radar Imaging is based on an extension of existing monostatic ISAR algorithms to the multistatic environment. The paper describes the ISAR processing and fusion by incoherent summation, and coherent summation, with and without overlapping observation angles. We demonstrate the proposed methods on simulated measurements.

A light weight SAR has been designed and built, suitable for use on short range tactical UAVs. The system consists of a fully digital receive array, and a very compact active transmit antenna. The approximate weight of the complete system is 6 kg, with power consumption below 75 W, depending on the required data acquisition volume. This X-band system is designed to provide an image resolution down to 10-15 cm at a maximum range of approximately 5 km. The design is described, and initial ground test results are presented.

High resolution radar imaging techniques can be used in ballistic missile defence systems to determine the type of ballistic missile during the boost phase (threat typing) and to discriminate different parts of a ballistic missile after the boost phase. The applied radar imaging technique is 2D Inverse Synthetic Aperture Radar (2D-ISAR) in which the Doppler shifts of various parts of the ballistic missile are employed to obtain a high cross-range resolution while the resolution in downrange is achieved with a large radar bandwidth. For a 10 cm downrange resolution, a radar bandwidth of more than 1.5 GHz is required. However, this requirement is not compatible with EM frequency spectrum allocations for long range ballistic missile defence radars that operate in the L, S, and C frequency band. In this paper, a novel coherent multiband ISAR imaging technique is proposed that employs two or more narrowband radar systems that operate in different frequency bands. The coherent multiband imaging process uses an advanced interpolation technique to achieve a very high downrange resolution and produces little artifacts due to noise.

Human motion estimation is an important issue in automotive, security or home automation applications. Radar systems are well suited for this because they are robust, are independent of day or night conditions and have accurate range and speed domain. The human response in a radar range-speed-time measurement behaves like an extended target where legs and arms coincide. A mutually non-coherent radar sensor network makes it possible to estimate additional information of the extended target response. To keep the system low cost the network uses commercial off-the-shelf (COTS) radar sensors without synchronisation of frequency or phase between the radar sensors. This article presents the results of human motion estimation with a mutually non-coherent radar sensor network. The calibration, radar processing, parameter estimation and classification of extended human objects are described. The swinging and rotating moving body parts give elliptical shapes in the differential range-speed responses. A model fit gives the legs and arms parameters on which classification is possible.

Radar can be an extremely useful sensing technique to observe persons. It perceives persons behind walls or at great distances and in situations where persons have no or poor visibility. Human motion modulates the radar signal which can be observed in the spectrogram of the received signal. Extraction of these movements enables the animation of a person in virtual reality. The authors focus on a fast feature-based approach to estimate human motion features for real-time applications. The human walking model of Boulic is used, which describe the human motion with three parameters. Personification information is obtained by estimating the individual leg and torso parameters. These motion parameters can be estimated from the temporal maximum, minimum and centre velocity of the human motion distribution. Three methods are presented to extract these velocities. Additionally, we extract an independent human motion repetition frequency estimate based on velocity slices in the spectrogram. Kalman filters smooth the parameters and estimate the global Boulic parameters. These estimated parameters are input to the human model of Boulic which forms the basis for animation. The methods are applied to real radar measurements. The animated person generated with the extracted parameters provides a realistic look-alike of the real motion of the person.

Radar observes targets, but they remain difficult to interpret due to the difficulty in analysing the radar range-speed sequences. The need for accurate analyses tools increases in case of extended target behaviour or multiple channel radars which give additional observation angles. Extended targets are targets with multiple scatterer responses which disturb each other and give a blurred target response. We investigate here the approach of deconvoluting the range-speed response with a point spread function and interpolate the range-speed positions to get the inner structure of the extended target. The deconvolution gives the individual elements of the extended target. The range-speed interpolation gives accurate position information. The positions and additional observation angle information are tracked with a filter. We demonstrate the approach with real radar measurements.

Radar can be used to observe humans that are obscured by objects such as walls. These humans cannot be visually observed. The radar measurements are used to animate an obscured human in virtual reality. This requires detailed information about the motion. The radar measurements give detailed information about the movements of the human body parts; the Doppler signatures are time-varying and observed in the spectrogram. The authors focus on the extraction of parameters and describe a method for estimating human walking parameters from radar measurements. The parameters are estimated by minimising the difference between a simulated model and real measurements. A human walking model is presented which can be used both to calculate the radar response and to visually animate a walking person. The method is applied to real radar measurements of inbound walking humans from a distance of 20 m. The results show that estimation of the walking parameters is possible. The animated walking human generated with estimated parameters is a realistic likeness of the real walking human.

Area surveillance for guarding and intruder detection with a combined camera radar sensor is considered. This specific sensor combination is attractive since complementary information is provided by the respective elements. Thus, a more complete description of objects of interest can be obtained. Several strategies to fuse the data are discussed. Results obtained with 'live' experiments are presented. When compared to camera only, a significant reduction of the number of false tracks is achieved.