An Efficient Preconditioner for RFICs Simulation Using Harmonic Balance Method

To overcome the limitation of Krylov subspace techniques using standard preconditioners in harmonic balance simulations of RFICs, a novel preconditioner, by virtue of a sparse-subsectional matrix, is introduced. This preconditioner, different from the ordinary block-diagonal matrix, not only gives a sufficient approximation to Jacobian matrix, but also exploits the characteristics of sparseness and subsectional condensation, which dramatically reduces memory usage and greatly improves factorization speed. Numeric experiments are included to demonstrate merits and efficiency of this new preconditioner