New Improved Sierpinski Carpet Structures Based Thinned Planar Array to Synthesize Low Sidelobes Radiation Pattern

This paper presents a methodology for generating new fractal structures characterized by its ability to reduce the number of elements in planar arrays while providing the desired beam figuration demand. The main challenges for creating fractal arrays such as the Sierpinski carpet are the high sidelobe level (SLL) as well as the large number of radiating elements in the higher growth stages. In this paper, based on the traditional Sierpinski carpet, several innovative fractal array structures are proposed which are embedded with traditional fully optimized planar arrays to reduce the computational and RF chain complexities in different growth stages and make them suitable for the application of global optimization techniques. Thus, it is possible to use the genetic algorithm (GA) to improve the excitation of elements in the proposed structures to generate the desired patterns. The method depends on generating different primitive array matrices with multiple possibilities consisting of working (ON) and non-working (OFF) elements, which is the basis for building improved planar arrays. To reduce RF complexity further, amplitude-only excitation is included in the optimization process. Computational results show that the new and improved versions of the Sierpinski carpet produce lower SLL radiation patterns with fewer radiating elements than their original counterpart in the fully optimized planar arrays.