ELEDIA Student Project Activities and Theses Reports: No conditions. Results ordered -Date Deposited. 2017-09-26T21:58:12ZEPrintshttp://eledia.science.unitn.it/images/eledialogo.pnghttp://eledia.science.unitn.it/publications/6972012017-09-22T15:04:34Z2017-09-22T15:04:34Zhttp://eledia.science.unitn.it/publications/id/eprint/748This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7482017-09-22T15:04:34ZA System‐by‐Design Approach for the Synthesis of WAIMs for Planar ArraysIn this work, wide-angle impedance matching (WAIM) layers based on metasurfaces are designed to enhance the radiation efficiency of planar phased arrays. Toward this end, a System-by-Design (SbD) approach is adopted where the layer geometrical features are the solution descriptors, while the minimization of the array power reflection is the synthesis objective. Selected numerical examples are shown to illustrate the potentialities of the proposed SbD-based design strategy.G. OliveriM. SalucciN. AnselmiA. Massa2017-09-15T11:56:56Z2017-09-15T11:56:56Zhttp://eledia.science.unitn.it/publications/id/eprint/747This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7472017-09-15T11:56:56ZInnovative Synthesis of WAIM Layers for Waveguide‐Fed Planar Phased ArraysThis work deals with the design of wide-angle impedance matching (WAIM) layers aimed at mitigating reflection issues arising in waveguide-fed planar phased arrays. The synthesis problem is formulated within the System-by-Design (SbD) framework, by minimizing the antenna input reflections caused by impedance mismatching when the array is steered through the optimization of the geometrical descriptors of the WAIM unit cells. Some numerical results are shown in order to assess the effectiveness of the proposed synthesis strategy.G. OliveriM. SalucciN. AnselmiA. Massa2017-09-08T14:26:08Z2017-09-08T14:26:08Zhttp://eledia.science.unitn.it/publications/id/eprint/746This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7462017-09-08T14:26:08ZSystem‐by‐Design Synthesis of Wide Angle Impedance Matching LayersIn this work, the synthesis of metasurface-based wide-angle impedance matching (WAIM) layers is proposed. The designed structures allow to mitigate reflection issues in waveguide-fed planar phased arrays. To achieve such a goal, the synthesis problem is formulated in the System-by-Design (SbD) framework, and the degrees of freedom are represented by the geometrical descriptors of the metasurface unit cells. The final layout is then obtained by minimizing the antenna input reflections caused by impedance mismatching when the array is steered. A preliminary numerical validation is provided in order to assess the effectiveness and flexibility of the proposed SbD-based design approach.G. OliveriM. SalucciN. AnselmiA. Massa2017-08-23T14:52:46Z2017-08-23T14:52:46Zhttp://eledia.science.unitn.it/publications/id/eprint/745This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7452017-08-23T14:52:46ZDictionary‐Based Bayesian Compressive Sensing for Imaging Arbitrary ScatterersThis work deals with an innovative free-space inverse scattering technique. The developed methodology is based on the exploitation of a Bayesian Compressive Sensing (BCS) solver and a set (or dictionary) of expansion bases. Several BCS-regularized reconstructions are performed using the different bases in the dictionary, and the sparsest solution is selected as the most reliable one. Thanks to such an approach, (i) no a-priori information about the unknown scatterers is required, and (ii) it is possible to extend the range of applicability of standard BCS-based inversion to objects having arbitrary size and shape. In order to verify the effectiveness of the proposed technique, as well as to test its robustness to noise, some illustrative numerical results are shown in the following. N. AnselmiG. OliveriM. HannanM. SalucciA. Massa2017-08-14T12:40:26Z2017-08-14T12:40:26Zhttp://eledia.science.unitn.it/publications/id/eprint/744This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7442017-08-14T12:40:26ZAn Innovative BCS‐Based Microwave Imaging Technique for Imaging Unknown Objects With Arbitrary Size and ShapeThis work presents a numerical validation of an innovative two-dimensional (2D) microwave inverse scattering technique exploiting Bayesian Compressive Sensing (BCS) and a dictionary of wavelet-based expansion bases. The goal of the dictionary-based BCS is to provide faithful guesses of the dielectric distribution inside the imaged scenario even if the unknown objects inside it are not sparse in the standard pixel basis. The developed strategy is based on a two-level hierarchical application of the BCS algorithm. In the first step, several sparsity-regularized inversions are performed using the dictionary of candidate bases. In the second step, the retrieved vectors are compared and the sparsest reconstruction is selected. Some numerical results are shown, in order to verify the effectiveness of the developed microwave imaging technique. Moreover, some illustrative results are shown to compare its performance with respect to competitive state-of-the-art alternatives.N. AnselmiG. OliveriM. HannanM. SalucciA. Massa2017-08-04T15:19:45Z2017-08-04T15:19:45Zhttp://eledia.science.unitn.it/publications/id/eprint/743This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7432017-08-04T15:19:45ZExtending the Applicability Range of Compressive Sensing‐Based Microwave Imaging to Arbitrary Scatterers This work deals with an innovative two-dimensional (2D) free-space microwave imaging technique. The developed inverse scattering (IS) technique is aimed at enabling Compressive Sensing (CS) to deal with the retrieval of unknown scatterers which are not necessarily sparse in the standard sense, i.e., in the pixel domain. Accordingly, the proposed technique exploits a user-defined dictionary of expansion bases that are used to retrieve several guesses of the electromagnetic properties of the investigation domain. Then, following the BCS paradigm, the sparsest solution is recognized as the optimal one. Some numerical results are presented, in order to verify the effectiveness of the proposed IS technique for imaging scatterers with arbitrary size and shape.N. AnselmiG. OliveriM. HannanM. SalucciA. Massa2017-07-28T15:46:06Z2017-07-28T15:46:06Zhttp://eledia.science.unitn.it/publications/id/eprint/742This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7422017-07-28T15:46:06ZInnovative Alphabet‐Based Bayesian Compressive Sensing Technique for Imaging Targets with Arbitrary ShapeIn this work an innovative two-dimensional (2D) microwave imaging technique exploiting Bayesian Compressive Sensing (BCS) and a wavelet-based alphabet for representing the problem unknowns is dealt with. The proposed approach is based on the generalization of the sparsity concept, extending the range of applicability of BCS-based inverse scattering (IS) techniques to objects with arbitrary shape and dimensions. A set of BCS reconstructions is performed considering different expansion bases in the alphabet, without the need for a-priori knowledge about the unknown scatterers. Then, the best reconstruction is recognized as that minimizing the number of non-null retrieved coefficients (i.e., the sparsest one). In order to verify the effectiveness of the proposed imaging technique, a set of representative numerical benchmarks is presented. Some comparisons with state-of-the-art IS techniques are presented, as well.N. AnselmiG. OliveriM. HannanM. SalucciA. Massa2017-07-21T12:47:48Z2017-07-21T13:45:32Zhttp://eledia.science.unitn.it/publications/id/eprint/741This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7412017-07-21T12:47:48ZFree‐Space Microwave Imaging through Alphabet‐Based Bayesian Compressive SensingA key requirement to be satisfied when exploiting Compressive Sensing (CS) methods in inverse scattering (IS) problems is that the unknowns (e.g., the contrast function or the equivalent sources) are sparse with respect to the considered expansion basis. State-of-the-art CS-based microwave imaging techniques typically consider single-resolution pixel-based representations, limiting their domain of applicability to the retrieval of few and isolated pixels within the investigated domain. Within this framework, this work is aimed at extending the range of applicability of CS-based approaches to the retrieval of unknown scatterers having arbitrary shape and dimensions. Since in real applications no a-priori information about the investigation domain is available, the idea is to retrieve a set of "candidate" solutions by executing several CS inversions using different expansion bases (e.g., pixel, Haar wavelets, Meyer wavelets, ...). Following the CS paradigm, the "best" solution can then be identified as the sparsest one, i.e., the solution with the lowest number of non-zero retrieved coefficients. A preliminary numerical validation of the proposed alphabet-based CS microwave imaging technique is given. Some numerical comparisons with competitive state-of-the-art inverse scattering techniques is shown, as well.N. AnselmiG. OliveriM. HannanM. SalucciA. Massa2017-07-14T11:17:21Z2017-07-14T11:17:21Zhttp://eledia.science.unitn.it/publications/id/eprint/740This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7402017-07-14T11:17:21ZAn Evolutionary Optimization Method for Solving the Multi-Frequency GPR Subsurface ProblemIn this work, the two-dimensional subsurface imaging problem is solved by processing wide-band ground penetrating radar (GPR) data. Towards this end, an innovative microwave inverse scattering technique is presented. The developed methodology is based on a particle swarm optimization (PSO) solver to minimize the arising multi-frequency (MF) cost function measuring the mismatch between measured and retrieved data. Moreover, an iterative multi-resolution strategy is exploited in order to progressively and adaptively refine the resolution of the retrieved images only within the regions of interest in which the presence of a buried object has been detected. Some numerical results are shown, in order to assess the effectiveness of the developed MF-IMSA-PSO strategy in reconstructing dielectric objects buried at different depths. Some experiments are also shown to verify the robustness of the proposed method when the background permittivity is not constant but smoothly varying with the distance from the interface.M. SalucciL. PoliN. AnselmiA. Massa2017-07-11T16:00:24Z2017-07-12T14:06:59Zhttp://eledia.science.unitn.it/publications/id/eprint/739This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7392017-07-11T16:00:24ZMulti‐Resolution Processing of Multi‐Frequency GPR Data for Robust Buried Object ImagingThis work presents an innovative GPR microwave imaging technique aimed at
retrieving the electromagnetic properties of inaccessible domains buried below a
planar interface. The arising two‐dimensional (2‐D) inverse scattering problem is
solved taking into account for the wide‐band nature of GPR data by exploiting a
multi frequency (MF) solution approach. Moreover, a customized multiresolution
particle swarm optimizer (IMSA‐PSO) is exploited in order to minimize
the MF cost function by adaptively refining the image resolution only in the
identified regions of interest (RoIs). A set of numerical experiments is shown in
order to verify the effectiveness of the developed MF‐IMSA‐PSO technique when
the background permittivity is not exactly known. A comparative assessment
with respect to a deterministic local search‐based microwave imaging technique
is given, as well, to highlight the superior performances yielded by the
exploitation of the PSO solver.M. SalucciL. PoliN. AnselmiAndrea Massa2017-06-29T07:27:29Z2017-06-29T07:27:29Zhttp://eledia.science.unitn.it/publications/id/eprint/737This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7372017-06-29T07:27:29ZRobust Multi-Frequency GPR Microwave Imaging through Multi-Scaling Particle Swarm OptimizationIn this work, an innovative GPR microwave imaging technique is proposed for solving the subsurface inverse scattering problem in a multi-frequency (MF) framework. The proposed technique exploits a stochastic optimizer based on particle swarm optimization (PSO) in order to effectively deal with the minimization of the MF cost function without being trapped into false solutions. Moreover, the iterative multi-scaling approach (IMSA) is exploited in order to achieve higher resolutions within the identified regions of interest. Some numerical results are shown, carefully selected from an extensive validation of the method, in order to assess its performance when dealing with the retrieval of buried objects under several noise levels. More precisely, a variation of the number of measurement points placed above the interface to collect the scattered radargram is considered, in order to investigate the robustness of the developed method, as well as to compare it to a deterministic implementation within the same solution framework.M. SalucciL. PoliN. AnselmiA. Massa2017-06-22T07:20:17Z2017-06-22T07:20:17Zhttp://eledia.science.unitn.it/publications/id/eprint/736This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7362017-06-22T07:20:17ZAn Innovative Multi-Frequency PSO-Based Method for the Microwave Imaging of
Buried Objects having Different Conductivities In this work, an innovative particle swarm optimization (PSO)-based microwave imaging approach is presented to solve the subsurface inverse scattering problem. The proposed MF-IMSA-PSO method integrates a customized PSO solver within a multi-scaling technique (i.e., the IMSA) in order to limit the ratio between problem unknowns and non-redundant data, mitigating the negative effects of both non-linearity and ill-posedness through the exploitation of progressively acquired information about the solution. Moreover, the inversion is performed by considering a multi-frequency (MF) solution strategy, by jointly processing several frequency components extracted from the spectrum of the measured data through ground penetrating radar (GPR). Some numerical results are shown in order to verify the effectiveness of the developed GPR microwave imaging technique when dealing with objects having a conductivity different from that of the hosting (lossy) soil.M. SalucciL. PoliN. AnselmiA. Massa2017-06-15T07:14:20Z2017-06-15T07:14:20Zhttp://eledia.science.unitn.it/publications/id/eprint/735This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7352017-06-15T07:14:20ZMicrowave Imaging of Buried Objects having Different Permittivities through an Innovative Multi-Frequency Stochastic MethodThis work deals with the retrieval of the electromagnetic characteristics of inaccessible subsurface domains by processing ground penetrating radar (GPR) data. Assuming a multi-frequency (MF) formulation of the buried inverse scattering problem, the solution is obtained by means of a multi-resolution particle swarm optimization (PSO) algorithm. The developed MF-IMSA-PSO method is able to proficiently exploit the intrinsic frequency diversity of wideband GPR measurements in order to mitigate the ill-posedness and non-linearity issues of the subsurface inverse scattering problem. Moreover, thanks to the integration of the PSO within the iterative multi-scaling approach (IMSA) an increased resolution of the retrieved images is obtained within the identified regions of interest, where the buried objects are supposed to lie. Some numerical experiments are shown in order to assess the effectiveness, the robustness to noise, as well as the current limitations, of the developed method in retrieving buried scatterers having different levels of electric permittivity (i.e., different levels of contrast with respect to the surrounding background medium). Moreover, a direct comparison with respect to the MF-IMSA-CG, a state-of-the-art approach based on a conjugate gradient (CG) local search algorithm, is given.M. SalucciL. PoliN. AnselmiA. Massa2017-06-08T13:16:17Z2017-06-08T13:16:17Zhttp://eledia.science.unitn.it/publications/id/eprint/734This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7342017-06-08T13:16:17ZRobust Multi-Frequency Subsurface Imaging through Evolutionary OptimizationIn this work, an innovative stochastic method for subsurface microwave imaging is presented. The proposed approach solves the subsurface inverse scattering problem by jointly processing multiple frequency components of the measured wide-band ground penetrating radar (GPR) data. Moreover, an iterative multi-zooming approach is adopted, in order to reduce the ratio between problem unknowns and informative data, as well as to adaptively enforce increased resolutions in correspondence with the identified regions of interest. The minimization of the multi-frequency (MF) cost function is performed at each multi-resolution step by means of a customized particle swarm optimization (PSO) algorithm, thanks to its capability of escaping from local minima, corresponding to false solutions of the inverse scattering problem. Some numerical results are shown, in order to assess the performance of the developed MF-IMSA-PSO method in retrieving buried targets having different shape and composition, as well as to compare it to a deterministic implementation within the same framework (i.e., the MF-IMSA-CG).M. SalucciL. PoliN. AnselmiA. Massa2017-06-01T10:10:22Z2017-06-01T10:10:23Zhttp://eledia.science.unitn.it/publications/id/eprint/733This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7332017-06-01T10:10:22ZAn Innovative Particle Swarm Optimization‐Based Approach for GPR Microwave ImagingThis work presents an innovative microwave imaging technique for accurate and robust subsurface imaging. The proposed approach is based on the integration of a customized particle swarm optimization (PSO) algorithm within the iterative multi-scaling approach (IMSA), and exploits multiple frequency components extracted from ground penetrating radar (GPR) wideband data. The solution of the arising inverse scattering problem is yielded within a multi-frequency (MF) approach, allowing to exploit the intrinsic frequency diversity of GPR measurements in order to add information and mitigate the ill-posedness and non-linearity issues. Some numerical experiments are shown in order to assess the effectiveness of the proposed MF-IMSA-PSO method when dealing with the retrieval of unknown buried scatterers having different shape. Moreover, a comparison to a competitive state-of-the-art deterministic approach is shown, in order to highlight the benefits of exploiting a global optimization algorithm in minimizing the MF cost function.M. SalucciL. PoliN. AnselmiA. Massa2017-05-25T07:22:11Z2017-05-25T07:22:11Zhttp://eledia.science.unitn.it/publications/id/eprint/732This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7322017-05-25T07:22:11ZMulti-Frequency Multi-Resolution Stochastic Optimization for GPR Microwave ImagingIn this work, the retrieval of the dielectric characteristics of unknown objects buried in a lossy half-space is dealt with. An innovative multi-resolution multi-frequency (MF) stochastic microwave imaging technique is proposed to solve the buried inverse scattering problem by processing wide-band ground penetrating radar (GPR) data. The proposed MF-IMSA-PSO method exploits a particle swarm optimization (PSO)-based algorithm to find the global optimum of the MF cost function measuring the mismatch between available and retrieved data at a fixed set of frequencies. Such a stochastic solver is nested within the iterative multi-scaling approach (IMSA) in order to reduce the ratio between problem unknowns and informative data, as well as to adaptively enforce increasing resolutions only within the regions of interest in which the scatterers have been detected. A preliminary numerical validation is shown, in order to assess the robustness of the developed approach with respect to noise, as well as to compare its performance to state-of-the-art competitive approaches.M. SalucciL. PoliN. AnselmiA. Massa2017-05-19T14:36:17Z2017-05-19T14:38:25Zhttp://eledia.science.unitn.it/publications/id/eprint/731This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7312017-05-19T14:36:17ZFrequency‐Hopping Multi‐Resolution GPR Microwave Imaging Based on Stochastic OptimizationThis work presents an innovative microwave imaging technique for processing wide-band ground penetrating radar (GPR) data and solving the subsurface inverse scattering problem. The proposed technique is based on the integration of a customized stochastic solver based on the Particle Swarm Optimizer (PSO) with the iterative multi-scaling approach (IMSA). Moreover, the IMSA-PSO is nested within a frequency-hopping (FH) approach in order to exploit the frequency-diversity of GPR measurements as an additional source of information for regularizing the subsurface inverse scattering problem. Some preliminary numerical results are shown in order to assess the effectiveness of the proposed methodology, as well as to compare it to a single-resolution (BARE) implementation within the same framework.M. SalucciL. PoliN. AnselmiA. Massa2017-05-04T06:46:03Z2017-05-04T06:46:03Zhttp://eledia.science.unitn.it/publications/id/eprint/730This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7302017-05-04T06:46:03ZMicrowave Imaging of Buried Targets through a Multi-Zooming Approach: Reconstruction Capabilities for Different Object ConductivitiesIn this work, the performance of an innovative microwave imaging methodology for buried object detection are analyzed. More precisely, the developed inverse scattering (IS) approach is based on a Multi-Frequency (MF) formulation of the buried IS equations in order to exploit the frequency diversity coming from wideband ground penetrating radar (GPR) measurements. The arising MF cost function is minimized through a customized deterministic solver based on a conjugate gradient (CG) minimizer nested within the iterative multi-scaling approach (IMSA) for achieving higher resolutions in the identified regions of interest (RoIs). Some illustrative numerical results are shown, in order to verify the effectiveness of the developed MF-IMSA-CG methodology when dealing with the retrieval of buried objects having different values of electric conductivity. For completeness, as well as for the sake of comparison, the reconstructions yielded by a competitive state-of-the-art approach based on a frequency hopping (FH) processing of the GPR spectrum are also shown, by considering several noise conditions.M. SalucciL. PoliA. Massa2017-04-27T07:14:23Z2017-04-27T07:14:23Zhttp://eledia.science.unitn.it/publications/id/eprint/729This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7292017-04-27T07:14:23ZMulti-Frequency Deterministic Inversion of Wideband GPR Data: Achievable Performance in the Presence of Unknown Background MediaThis work deals with the retrieval of the electromagnetic properties of buried investigation domains starting from the processing of wideband ground penetrating radar (GPR) data. More precisely, the achievable performance by two deterministic multi-zooming (IMSA) conjugate gradient (CG)-based techniques are assessed when no exact knowledge of the electric permittivity of the background is available to perform the inversion. More precisely, the first analyzed technique is based on a multi-frequency (MF) approach (i.e., the MF-IMSA-CG), while the second one is a state-of-the-art frequency-hopping (FH) based methodology (i.e., the FH-IMSA-CG). Some significant numerical results are reported in order to understand what is the impact of a wrong guess of the hosting medium under several noise conditions for both MF and FH-based inversion methods.M. SalucciL. PoliA. Massa2017-04-20T07:14:15Z2017-04-20T07:14:15Zhttp://eledia.science.unitn.it/publications/id/eprint/728This item is in the repository with the URL: http://eledia.science.unitn.it/publications/id/eprint/7282017-04-20T07:14:15ZDeterministic Inversion of Wideband GPR Data for Multi‐Resolution Imaging of Buried ObjectsIn this work, the performance of an innovative conjugate-gradient (CG)-based deterministic microwave imaging technique are assessed for the inversion of wideband ground penetrating radar (GPR) data. The developed methodology exploits the intrinsic frequency diversity of GPR measurements through a multi-frequency (MF) strategy in order to add information to the inversion problem and to mitigate the negative effects of ill-posedness of the buried inverse scattering (IS) problem. Moreover, the iterative multi-scaling approach (IMSA) is exploited in order to increase as much as possible the ratio between non-redundant data and problem unknowns, thus mitigating the problem of the non-linearity. Some numerical results are shown, in order to analyze the achievable reconstruction capabilities by the developed MF technique when dealing with the retrieval of objects having different values of relative permittivity. A direct comparison with a frequency hopping (FH)-based implementation of the same multi-resolution deterministic solver is shown, as well, to highlight the differences between the two approaches.M. SalucciL. PoliA. Massa