Jesús A. López-Fernández

Fast Antenna Characterization Using the Sources Reconstruction Method on Graphics Processors

The Sources Reconstruction Method (SRM) is a noninvasive technique for, among other applications, antenna characterization. The SRM is based on obtaining a distribution of equivalent currents that radiate the same field as the antenna under test. The computation of these currents requires solving a linear system, usually ill-posed, that may be very computationally demanding for commercial antennas. Graphics Processing Units (GPUs) are an interesting hardware choice for solving compute-bound problems that are prone to parallelism. In this paper, we present an implementation on GPUs of the SRM applied to antenna characterization that is based on a compute-bound algorithm with a high degree of parallelism. The GPU implementation introduced in this work provides a dramatic reduction on the time cost compared to our CPU implementation and, in addition, keeps the low-memory footprint of the latter. For the sake of illustration, the equivalent currents are obtained on a base station antenna array and a helix antenna working at practical frequencies. Quasi real-time results are obtained on a desktop workstation.

Acoustic scattering solver based on single level FMM for multi-GPU systems

In this paper, we present a heterogeneous parallel solver of a high frequency single level Fast Multipole Method (FMM) for the Helmholtz equation applied to acoustic scattering. The developed solution uses multiple GPUs to tackle the compute bound steps of the FMM (aggregation, disaggregation, and near interactions) while the CPU handles a memory bound step (translation) using OpenMP. The proposed solver performance is measured on a workstation with two GPUs (NVIDIA GTX 480) and is compared with that of a distributed memory solver run on a cluster of 32 nodes (HP BL465c) with an Infiniband network. Some energy efficiency results are also presented in this work.

Optimization framework on antenna arrays for near field multifocusing

This paper presents a new method for focusing on multiple targets simultaneously using antenna arrays. An optimization framework based on the Levenberg-Marquardt algorithm is used to concentrate the radiated field at certain positions. The proposed synthesis technique is a flexible method that can optimize different sets of parameters in the array, i.e., the weights (phase and/or magnitude) of the elements, their positions or all these features, in order to achieve this goal. Some illustrative examples using simulated and implemented antennas are addressed to evaluate this framework.

A GPGPU solution of the FMM near interactions for acoustic scattering problems

The Fast Multipole Method (FMM) is specially suitable for applications in which it is necessary to predict the acoustic scattering, e.g., aircraft noise control. This accelerated iterative method has two main parts, far interactions and near interactions. Near interactions are computationally intensive and they fit properly in the Single Instruction Multiple Threads paradigm. In this work, we present a heterogeneous parallel solution in which the near interactions are computed using Graphical Processing Units (GPUs). The performance of the proposed solution is proved using a workstation with one NVIDIA GTX480 GPU and a cluster that consists of 32 nodes HP BL465c with an Infiniband network.

A MULTI-GPU SOURCES RECONSTRUCTION METHOD FOR IMAGING APPLICATIONS

A proflle reconstruction method using a surface inverse currents technique implemented on GPU is presented. The method makes use of the internal flelds radiated by an equivalent currents distribution retrieved from scattered fleld information that is collected from multiple incident flelds. Its main advantage over other inverse source-based techniques is the use of surface formulation for the inverse problem, which reduces the problem dimensionality thus decreasing the computational cost. In addition, the GPU implementation drastically reduces the calculation time, enabling the development of real time and accurate geometry reconstruction at a low cost.

Echo Characterization Based on Maximum-Likelihood Estimation for Antenna-Measurement Correction [Measurements Corner]

Antenna measurement systems could be affected by multipath effects resulting in reflected beams that might modify the measured values. In this paper, the use of the Maximum Likelihood estimator provides the basis of several post-processing methods able to estimate the echo parameters in imperfect antenna measurements. Once the reflections are characterized, these parameters are used to correct the measured values and to retrieve the actual radiation pattern corresponding to the antenna under test. Some different techniques are proposed in order to perform such cancellation. Experimental validation is also presented to show the accuracy of the described methods.

Fast and Accurate Modeling of Dual-Polarized Reflectarray Unit Cells Using Support Vector Machines

This paper describes the characterization of reflectarray unit cells using support vector machines (SVMs) to obtain fast and accurately the full matrix of reflection coefficients, which is used for an analysis of dual-polarized reflectarrays, demonstrating the performance of the model. First, a surrogate model of the reflectarray unit cell is obtained using SVMs. To this end, a set of random samples of the reflection coefficient matrix with a full-wave method of moments based on local periodicity (MoM-LP) is used to train the SVMs. To efficiently obtain the surrogate model, a novel strategy to accelerate the training process is presented, remarkably reducing computing time. Next, the model is tested against a different set of samples, obtaining an excellent agreement between the SVM model and MoM-LP simulations for all reflection coefficients, including the cross-coefficients. The surrogate model is then used for an efficient analysis of three reflectarrays with pencil beam for point-to-point communications, isoflux pattern for global Earth coverage, and a shaped beam for local multipoint distribution service application, showing excellent agreement in both copolar and crosspolar patterns between the SVM and MoM-LP simulations. Finally, the analysis is accelerated by a factor larger than three orders of magnitude using SVMs instead of MoM-LP.

Improving the FMM performance using optimal group size on heterogeneous system architectures

The fast multipole method (FMM) is commonly used to speed-up the time to solution of a wide diversity of N-body type problems. To use the FMM, the elements that constitute the geometry of the problem are clustered into groups of a given size (D) that may deeply vary the time to solution of the FMM. The optimal value of D, in the sense of minimizing the time cost, is unknown beforehand and it depends on several factors. Nevertheless, during the solver setup, it is possible to produce clusters of varying size D and to estimate the time to solution associated with each D. In this paper, we use octree structures to efficiently perform clustering and time cost estimation to find the optimal group size for the FMM implementations on a heterogeneous architecture. In addition, two different frameworks have been analyzed: single-level FMM and fast Fourier transform FMM (FMM-FFT). We found that the sensitivity of the time cost to the parameter D depends on factors, such as the problem size or the implementation of the FMM framework. Moreover, we observed that the time cost may be conspicuously reduced if a proper D is employed.

Self-Calibration for Fault or Obstacle Correction in Continually Rotating Array Antennas

A novel self-calibration scheme for rotating array antennas is proposed. It is based on the acquisition of some near fleld samples using a static probe providing information about the actual behavior of the antenna. If any error, fault or obstacle modifles the desired behavior, the weights applied to the feedings of the array elements are modifled so that speciflcations are fulfllled again. Additionally, coupling between the elements of the arrays is also accounted for. Difierent disciplines such as near fleld to far fleld transformation, antenna modeling, adaptive flltering or automatic learning are involved in this system. Some signiflcant results are also presented.