open menu

Alessandro Cuttin

PhD student, University of Trieste

Alessandro Cuttin holds a MEng degree in Communications Engineering, achieved at the University of Trieste, where he is currently attending  the PhD program in Information Engineering. He is also co-founder of PicoSaTs s.r.l., a space start-up incorporated in Area Science Park, focused on providing new communication technologies to the family of CubeSat nanosatellites.
His current research interests cover problems of multiple random access, array antenna design and optimization, space communications.


Enhancing deterministic synthesis of antenna arrays with stochastic methods
Sala Terra, Tue, 17/05/2016 - 14:10 - 14:30

Nowadays, society is relying more and more on wireless communications and related technologies.
Antennas, as one of the key elements of a communication system, are subject of widespread research.
There are circumstances when a single element antenna cannot meet the needs of a specific application or scenario. In such cases, instead of an overly complex radiating element, a set of simpler antennas can be used. This technique is commonly referred to array antenna.

Array antennas have the capabilities to meet certain radiation characteristics (like radiation pattern, gain, and polarization) that single element radiators would have never achieved. This happens thanks to the increase in degrees of freedom on which a designer can act; these are, for instance: the number of elements, their geometrical configuration, and their excitation.

In the synthesis of an array (that is: the process of establishing the array characteristics in order to obtain a desired performance) it is not unfrequent to encounter non linear problems. The power pattern synthesis with respect to the excitations or the location is just an example. This family of problems has been, and currently is, subject of research; to solve them both deterministic and stochastic methods have been devised.

The choice of a deterministic method over a stochastic one depends on many factors. Generally, the first one is more fast than the second, but its convergence is not always guaranteed. We discuss the case of the optimization of a geometrically uniform array, whose geometrical envelope is imposed by the application scenario. In this case, the radiating elements are uniformly distributed over a regular grid.

In order to obtain a certain directivity and sidelobe suppression, the only quantities we can act on are the phase and the amplitude of the excitations. When a problem of this type is solved by means of stochastic methods, the main issue is the time required to obtain convergence. Instead, a deterministic method, like the successive projections method, is relatively fast in finding the solution. 

However, it relies on the definition of suitable masks, designed according to a given set of specifications and constraints. These can be satisfied by a number of different masks, not only one, and the successive projections method is very sensitive to their definition. We investigate the case where the deterministic method of successive projections is coupled with, and takes advantage of a stochastic one, in order to find the better mask that satisfies the requirements. In this way, given a certain array geometry, it can be established which are its top performances.

Sign in