Abstract: A new algorithm is presented for analyzing the RCS of large-scale finite arrays, using SED(sub-entire-domain functions) to construct the basis functions, FBM(Forward-Backward-Method) in solving the linear equations and DFT(Discrete Fourier Transform) to accelerate the inner product of the vector. Compared with the existing MLFMM which has been widely used, the new method which can guarantee the accuracy of results reduces the computational complexity and the memory storage from O(NlogN) to O(N)， where N is the number of unknowns of the array. This new method can be used to analyze very large finite arrays, and numerical examples are presented to demonstrate its validity.

Key words: method of moments(MOM), radar cross section(RCS), discrete fourier transform(DFT), sub-entire-domain functions (SED), forward-backward-method (FBM)