2,845 research outputs found
Detection and motion parameters estimation techniques in Forward Scatter Radar
No full textForward scatter Radar systems designed to take advantage of the greater radar cross section, that is robust to Radar Absorbing Material and other stealth technology, and of the long integration times, due to the little phase and amplitude fluctuations, are attractive for a variety of applications. Many of which fit well with the needs of augmentation of the surveillance capabilities of low-observable targets that may have a small backscatter RCS when observed with the conventional radar systems. This thesis reports on research into this field of radar systems with additional contributions to target detection and motion parameters estimation.
Particularly, the first part of the thesis deals with the detection of moving targets that follow a linear trajectory in a single node FSR configuration. The detection scheme based on a square-law detector followed by an appropriate matched filter, here addressed as Crystal Video Detector (CVD) following the traditional terminology (Crystal Video Receiver), has already been put forward in the literature. Performance prediction and FSR system design were key motivator to analytically characterize the detection performance of CVD in terms of both, probability of false alarm and probability of detection. The derived closed-form expressions were validate from Monte Carlo simulations under different geometrical conditions and from experimental data acquired by a passive FSR based on FM signals. Furthermore, new detection schemes based on the CVD ensuring the constant false alarm rate (CFAR) condition were devised and analytically characterized. The performance analysis showed quite small losses of the CFAR-CVD detectors compared to the fixed threshold CVD.
The second part of the thesis still handles the problem of target detection through the derivation of innovative detection schemes based on the Generalized Likelihood Ratio Test (GLRT). A comparison with the detection performance of the CVD has proven the better performance of the GLRT-based detectors. In most cases the improvement has an upper bound of 3 dB. However, there are specific circumstances where the standard FSR detector shows significant losses while the GLRT schemes suffer a much smaller degradation. Moreover the possibility to have a set of secondary data assumed target free, drove to the devising of new GLRT schemes. The results demonstrated a non-negligible further improvement over the previous GLRT schemes when the operation conditions get close to the near field transition point. The detection performance of the derived detectors without and with secondary data were analytically characterized. This analytical performance allowed to derive simplified equivalent SNR expressions that relate the GLRT detection performance to the main system and target parameters. These expressions showed to be useful for the design of effective FSR geometries that guarantee desired detection performance for specific targets.
In the third part of the thesis the focus is moved to the motion parameters estimation through both, a single baseline and a dual baseline FSR configuration. Accordingly, the Doppler signature extracted from the Crystal Video based scheme is exploited. Following motion parameters estimation approaches already introduced in the literature, a two dimensional filter bank technique was proposed. The main target parameters encoding Doppler rate, main lobe width and crossing time instant were estimated from such technique. The accuracy of the proposed technique was investigated from a theoretical point of view through the derivation of simplified closed-form expression of the Cramer Rao Lower Bound (CRLB). The analysis proved that unbiased estimates of the desired target parameters can be obtained that approach the derived CRLB in the high SNR region. After the dependence of the kinematic parameters on the parameters estimated from the bank was exploited. The cross baseline velocity in a single baseline configuration was estimated under the assumption that the baseline crossing point is known. Meanwhile the dual baseline configuration ensures the possibility to estimate also the baseline crossing point without a priori knowledge on the other target kinematic parameters. Once more, the CRLB of the target motion parameters for both reference scenarios was derived. The analysis proved that unbiased estimates of the target motion parameters can be obtained with high accuracy even for low SNR conditions. The effectiveness of the proposed approach was also shown from experimental data acquired by a passive FSR based on FM signals
Earth Observation with SAR Satellite Formations: New Techniques and Innovative Products
No full textThis paper discusses the use of small satellites for future radar remote sensing applications. After a short introduction, we give first an overview of the TanDEM-X mission to be launched in autumn 2009. Here, special emphasis is put on the demonstration of innovative synthetic aperture radar (SAR) imaging techniques. Then, novel SAR configurations are introduced which make synergistic use of multiple small satellites flying in close formation. Performance examples demonstrate their unique capabilities for advanced Earth observation
applications. Among these opportunities are the generation of digital elevation models with decimetre accuracy, the monitoring of ocean currents, and the measurement of small vertical displacements from snow accumulation, vegetation growth, and thawing permafrost soils. Challenges associated with the use of small satellites are pointed out and solutions to overcome them are presented
Spatial resolution improvement in GNSS-based SAR using multistatic acquisitions and feature extraction
No full textThis paper considers the exploitation of navigation satellite systems as opportunity transmitters for bistatic and multistatic synthetic aperture radar (SAR). The simultaneous availability of multiple satellites over a scene of interest at different viewing angles allows multistatic SAR acquisitions using a single receiver on or near the ground. The resulting spatial diversity could be used to drastically improve image resolution or to enhance image information space. To exploit the availability of multiple satellites, two data fusion approaches are here considered. In the former, point features of the single images obtained from different perspectives are extracted and then combined, whereas in the latter, a multistatic image is first obtained by combining the single channel data at the image level and then the point features are extracted. This is achieved by considering ad hoc CLEAN-like techniques. These techniques have been developed on both the analytical and simulation levels and experimentally verified with real GNSS-based SAR imagery. The techniques described here are not limited to GNSS-based SAR but may be applied to any multistatic SAR system
Passive multi-perspective GNSS-based SAR using CLEAN technique: An experimental study
No full textThis paper presents an experimental study on multiple perspectives GNSS-based SAR. The large number of nav-igation satellites illuminating the same area from different view angles allows passive SAR imagery at multiple bistatic geometries. The multiple perspectives may result in different displacement of the dominant scattering centers composing the scene. By applying ad-hoc CLEAN technique to individual bistatic images, the target points can be identified and subsequently the information can be fused. The experimental study here carried out demonstrates the potential of this technology to enhance the image information space for persistent local area monitoring purposes
Passive radar imaging of ship targets with GNSS signals of opportunity
No full textThis article explores the possibility to exploit global navigation satellite systems (GNSS) signals to obtain radar imagery of ships. This is a new application area for the GNSS remote sensing, which adds to a rich line of research about the alternative utilization of navigation satellites for remote sensing purposes, which currently includes reflectometry, passive radar, and synthetic aperture radar (SAR) systems. In the field of short-range maritime surveillance, GNSS-based passive radar has already proven to detect and localize ship targets of interest. The possibility to obtain meaningful radar images of observed vessels would represent an additional benefit, opening the doors to noncooperative ship classification capability with this technology. To this purpose, a proper processing chain is here conceived and developed, able to achieve well-focused images of ships while maximizing their signal-to-background ratio. Moreover, the scaling factors needed to map the backscatter energy in the range and cross-range domain are also analytically derived, enabling the estimation of the length of the target. The effectiveness of the proposed approach at obtaining radar images of ship targets and extracting relevant features is confirmed via an experimental campaign, comprising multiple Galileo satellites and a commercial ferry undergoing different kinds of motion
Hyperauthorship in Mikhail Bakhtin: The Primary Author and Conceptual Personae
No full textThis article explores the phemenon of hyperauthorship in intellectual writing: a primary author (hyperauthor) creates a number of secondary authors (hypoauthors), and develops possible conceptual systems on their behalf. The case under consideration is Mikhail Bakhtin and his complex relationship with his friends Pavel Medvedev and Valentin Voloshinov, members of the so called "Bakhtin's circle" (in the 1920s) who are credited with authorship of several books which may have been actually written by Bakhtin himself. Still unclear from biographical and historical perspectives, this problem of authentic attribution of Medvedev's and Voloshinov's texts can be clarified in the theoretical framework of "hyperauthorship" and "possibilistic thinking." This article applies Bakhtin's own theory of the "primary author immersed in silence," as well as Deleuze and Guattari's notion of "conceptual personae," to explain this case of "shared," or "transferred" authorship. The figures of Voloshinov and Medvedev, though historically real, may be viewed as Bakhtin's projections of "ideal," or "utopian" Marxism in linguistics and literary theory
24 GHz interferometric radar for road hump detections in front of a vehicle
No full textThis paper presents an interferometric radar system using SFCW to detect small road obstacles at 24 GHz. Experimental results confirm the effectiveness of the method employed to estimate the hump's height, and it can be used in the context of driver assistance or even driverless cars in a future
Sea clutter power reduction in pulse forward scatter radar
No full textA Forward Scatter Radar (FSR) system does not have range resolution so the clutter is collected from the whole surface illuminated by the antenna radar. This paper suggest an approach to reduce the sea clutter power, using a pulse Forward Scatter Radar. A resolution Shell is defined. Its introduction in target detection showed an improvement in Signal to Clutter Ratio (SCR)
Passive multi-static SAR with GNSS transmitters: First theoretical and experimental results with point targets
No full textThis paper investigates the potential in using a passive multi-static Synthetic Aperture Radar (SAR) with Global Navigation Satellite Systems (GNSS) as transmitters of opportunity. The motivation for this research is based on the concept that a non-coherent combination between multiple bistatic images, obtained by satellites with essentially different angular separations, may yield multi-static imagery with a drastically improved resolution in the range dimension. This concept is confirmed by analytically deriving the multi-static Point Spread Function (MPSF) for this topology. The validity of the analytical results is verified with experimental data. © VDE VERLAG GMBH · Berlin · Offenbach, Germany
What are the implications of Curriculum Learning strategy on IRL methods?: Investigating Inverse Reinforcement Learning from Human Behavior
No full textInverse Reinforcement Learning (IRL) is a subfield of Reinforcement Learning (RL) that focuses on recovering the reward function using expert demonstrations. In the field of IRL, Adversarial IRL (AIRL) is a promising algorithm that is postulated to recover non-linear rewards in environments with unknown dynamics. This study investigates the potential benefits of applying the Curriculum Learning (CL) strategy to the AIRL algorithm. For our experiments, we use a randomized partially observable Markov decision process in the form of a grid-world-like environment. Using only expert demonstrations obtained with an RL algorithm under the true reward function, we train AIRL in a variety of configurations and identify an effective curriculum. Our results show, that a well-constructed curriculum can enhance the performance of AIRL twofold in both key aspects: the speed of convergence and the efficiency of using expert demonstrations. We thus conclude that CL can be a useful addition to an AIRL-based solution. Full code is available online in the supplementary material https://github.com/mikhail-vlasenko/curriculum-learning-IRL.CSE3000 Research ProjectComputer Science and Engineerin
- …
