%X Monopulse radar applications require the realization of antennas able to generate two different patterns: a sum pattern and a difference pattern. In order to achieve this objective, antenna arrays are usually employed and each pattern is generated by using different and separated feeding networks. This solution allows one to obtain very good performances, but at the expense of a very high cost, due to the doubled complexity of the feeding network.
In order to overcome this disadvantage, a common approach consists in the reduction of the number of elements of the array, reducing then the complexity of the feeding network. This is possible by using arrays with a sparse geometry (i.e. arrays with non-equispaced elements): in this way the elements locations can be used in the design process as an additional degree of freedom, allowing the designer to satisfy the project requirements by using a lower number of elements with respect to the classical uniform array case.
By using two feeding networks it is possible to modify the pattern of the array by changing the excitation applied to each element, but the same cannot be done with the array positions. Hence, it is very important to define a set of element positions which is optimal for both patterns.
The objective of the activity is the development and the analysis of a technique based on Multi Task Bayesian Compressive Sensing for the synthesis of two sparse CONFORMAL arrays:
-with different pattern (sum an difference patterns)
-with the same position for each array element.

%D 2015
%A J. Moseley
%I University of Trento
%K Compressive Sensing, Array Synthesis, Direction-of-Arrival 
%L elediasc12670
%T Reconfigurable Conformal Array Synthesis through MT-BCS