Heterogeneous Array Pattern Approximation With Mutual Coupling

The patent divides the large-scale heterogeneous array into multiple sub-arrays, and further segments elements within sub-arrays into representative elements and ordinary elements. This segmentation allows full-wave simulations to be performed only on a small number of representative elements rather than all elements, dramatically reducing computational complexity while maintaining accuracy through the use of segmentation matrices to account for mutual coupling effects.

The patent creates simplified copies (representative elements) that capture the essential electromagnetic characteristics of groups of similar elements. These representative elements serve as templates whose patterns are then reused and adjusted through segmentation matrices to represent the behavior of many actual elements, avoiding the need to simulate each element individually.

The patent segments the array into sub-arrays and identifies representative elements within each sub-array. Full-wave simulations are performed only on these representative elements, generating a small set of pattern data. This segmentation approach reduces storage requirements from needing to store data for all hundreds or thousands of elements to storing data for only a handful of representative elements, while the segmentation matrices enable reconstruction of the complete array pattern.

The representative elements serve multiple functions: they capture the electromagnetic characteristics of their respective sub-arrays, and their patterns are reused across multiple elements through the segmentation matrix approach. This multi-functionality eliminates the need to store separate high-resolution pattern data for each individual element, significantly reducing storage overhead.

The patent combines the simplicity of pattern multiplication with the accuracy of mutual coupling consideration through segmentation. The array is divided into sub-arrays, and within each sub-array, representative elements are identified. The pattern multiplication theorem is applied to representative elements and their corresponding sub-arrays, while segmentation matrices are used to account for mutual coupling effects. This hybrid approach maintains computational simplicity while improving measurement precision.

The patent introduces segmentation matrices as intermediaries between the simple pattern multiplication approach and the accurate but complex full mutual coupling analysis. These matrices serve as a bridge, allowing the use of efficient pattern multiplication while incorporating mutual coupling effects through the matrix transformations, thus achieving both simplicity and accuracy.