Communication at Eucap 2016

Among other goals, our project pursues the development of scanners based on amplitude-only information acquired on a small set of points so that the overall cost is reduced.

However, both approximations are expected to introduce an error higher than the one observed in a conventional system. This cumulative error has been studied for the context of antennas measurements and the results have been presented in the European Conference on Antennas and Propagation hold in Davos, Switzerland (10-15 April).

You can find further details in the proceedings paper:

Ana Arboleya, Jaime Laviada, Juha  Ala Laurinaho, Yuri Álvarez, Fernando Las­-Heras and Antti V. Räisänen, “Reduced Set of Points in Phaseless Broadband Near­-Field Antenna Measurement: Effects of Noise and Mechanical Errors”, in Proc. European Conference on Antennas and Propagation, Davos, 10-15 April, 2016.

The paper will be also available later in ieeexplore.

Accurate characterization of broadband antennas by means of amplitude-only information

As it has been previously detailed in past entries, one of the goals of our project is to develop techniques to enable an electromagnetic scanner based on amplitude-only information (i.e., acquiring only the power of the signals). Nevertheless, it is also relevant these techniques are not only useful for electromagnetic imaging. In fact, there is great interest on these techniques in the antenna measurements area.

The accurate measurement of antennas at the millimeter and submillimeter-wave bands, which are currently subjected to really intense development activity, is very challenging due to the several difficulties. For example, these antennas are usually measured by employing frequency extension modules connected to a vector network analyzer. The frequency extension modules convert low frequency signals (usually up to 40 GHz) to millimeter- or submillimeter-wave bands (up to 1.5THz, but technology is quickly growing so this frequency limits are continuously broken).

The frequency extension modules are connected to the antenna under test and to the probe antenna by means of rigid waveguides. After that, the probe antenna (together with the extension module) is moved along a (typically) canonical surface such as a sphere, a cylinder or a plane to acquire the radiated field and, after some manipulations, obtain the radiation pattern of the antenna. You can check a nice video by NIST illustrating this process here.

Nevertheless, this movement ineludibly involves bending the cables which connect the network analyzer and the frequency extension modules. This bending usually has some impact in the phase of the signals. Although this impact is usually negligible at low frequencies, once these low frequency signals are translated to the millimeter- or submillimeter-wave regime, this impact becomes much more prominent yielding considerable inaccuracies. In addition, other phenomena such as temperature stability of the room can also have an impact on the accuracy of the phase measurement.

In our recent paper, which is based on the aforementioned techniques for electromagnetic imaging, we show how it is possible to perform accurate acquisitions by employing amplitude-only information. You can take a look to the technical details in the preview of the paper as well as check the final paper in ieeeplore.

Scheme for accurate broadband antenna characterization by means of amplitude-only information (phaseless). See further details in the paper.

Millimeter-wave through the wall vision from power signals

A new work has been published by the researchers of the project to illustrate the capabilities of electromagnetic imaging based on amplitude-only signals. As it has been described in this website, the development of scanners based on amplitude-only (i.e., phaseless) signals is one of the goals of the PORTEMVISION project as this kind of methods rely on simple acquisition schemes (scalar measurements), which are usually much more simple and, therefore, simpler (and cheaper!) to perform. Besides, these systems can easily (well, at least easier than those ones based on amplitude/phase acquisitions) work in broadband (i.e., for a large number of frequencies) and, in the electromagnetic world, it is translated into a better depth resolution.

The work presented in the paper describes how it is possible to compute images based on the electromagnetic signals scattered from a wall and the objects behind. The achieved images have excellent agreement with real-life images enabling the detection of metallic objects (e.g., pipes) at several tens of centimeters beyond the wall. Of course, this method does not damage the wall at all.

Object under test

Object behind plasterboard with the Ka band setup (ad-hoc components)

Object behind mortar and hollow bricks wall at th Ku band (commercial components).

Two setups have been implemented at different bands using commercial components (easy to reproduce!) as well as homemade components. Both systems have been tested for standard building walls: mortar, hollow bricks, plasterboard, wood, etc.

Three-dimensional image for the plasterboard setup.

Three-dimensional image for the hollow bricks and mortar setup.

You can find the preliminary version of the manuscript here. The final version will be available soon in ieeexplorer.

Published paper on scalar calibration

Everybody agrees that high quality images requires good resolution. If we consider 3D images, this resolution must be provided in the three cartesian axis.

In the electromagnetic context, high resolution is achieved by means of good lateral and depth resolutions. In the electromagnetic imaging context, lateral resolution is proportional to the wavelength so if you wish to achieve good resolutions you must move up in frequency. On the other hand, depth resolution (for instance, to know the distance between two points) is given by the bandwidth of the signal.

Due to technical reasons, both facts can be complexed/expensive to achieve. In our recently published paper, we provide a technique for electromagnetic imaging that only requires the intensity of the field to provide good images. In contrast to other techniques, the calibration of the system can be also done by using intensity data. For this reason, this enables to keep a simple architecture of the scanner without sacrificing resolution.

J. Laviada, F. Las-Heras, “Scalar Calibration for Broadband Synthetic Aperture Radar Operating with Amplitude-Only Data,” in IEEE Antennas and Wireless Propagation Letters, vol.14, no., pp.1714-1717, 2015.

 

You can find the manuscript here and the final edited copy (if you have the subscription) here.

Presentation at URSI 2015

Several members of the project are attending to the “XXX Simposium Nacional de la Unión Científica Internacional de Radio, URSI 2015” and the associated workshop on THz technology. In this context, one of the pillars of the project, the capacity of extract full information from an amplitude-only measurement, will be presented. In particular, the presented results correspond to antenna measurement context and they will be extrapolated to inverse scattering problems along the project.

Ana Arboleya Arboleya, Jaime Laviada Martínez, Juha Ala-Laurinaho, Yuri Álvarez López, Fernando Las-Heras, Antti Räisänen, “Extrapolation of Conventional Off-Axis Holography Method for Broadband Antenna Characterization”, URSI 2015, 2-4 sept. 2015

It is also relevant to remark that this work has been done in collaboration with members of the MilliLab.

It is also of interest to observe that members of the project have presented several previous works related to fast and reliable postprocessing of the data acquired by a multistatic Synthetic Aperture Radar (SAR) that will be useful for the on-going tasks:

1. Yolanda Rodríguez-Vaqueiro, , Yuri Álvarez, Borja González-Valdés, Fernando Las-Heras, Óscar Rubiños, Antonio García Pino, José Ángel Martínez Lorenzo, “A fourier-based Operator to Accelerate a Multistatic Compressive Sensing Imaging System”, URSI 2015, Sep. 2015.
2. Borja Gonzalez-Valdes, Yuri Álvarez, Yolanda Rodriguez-Vaqueiro, Fernando Las-Heras, Antonio García Pino, Jose A. Martinez Lorenzo, “Multistatic Backpropagation Imaging using a fourier-based Method”, URSI 2015, Sep. 2015.

Project has started!

Minister has given green light to start the projects after the 31st of July. Unfortunately, the host university closes from the 1st of August to the 17th of August… so it was not possible to activate the project so far. Anyway, today has been possible to officially start the project!

First things first, we are now working to build the framework to ease the execution of the tasks and working packages. Several providers are being contacted to find the best quality/price ratio and a strategy to unify the new equipment with the already available systems is under development.

Back to school in August!

Kick-off meeting

PORTEMVISION proposal has been approved in the framework of the Spanish R&D projects on the 22nd of April. Although the final resolution has not been published yet, the “proposal of final resolution” is available since the 26th of May and that means that the project is almost ready to start!.

For this reason, a kick-off meeting has been held between all the researchers of the project last friday. In the meeting, a general overview of the project has been presented and the discussion has been focused on working package schedule the ways to tackle the problem we are facing and equipment purchase.

Now, it is time to rest and wait for the final approval of the project.