An analytical derivation of the optimum source patterns for the pseudospectral time-domain method

In the computational electromagnetics and acoustics, spatially smoothed sources are often utilized to alleviate the aliasing errors in the pseudospectral time-domain (PSTD) algorithms. In our work, an analytical derivation of the optimum source patterns is presented according to the accurately derived expressions of the dominant source-introduced aliasing errors according to the circular discrete convolution and Tailor series expansion method. We quantitatively demonstrate, for the first time in literature, that the aliasing errors can be optimally suppressed and rapidly reduced to the negligible levels by these optimum patterns and with the increment of source cells. We also provide the different implementation schemes of the optimal patterns both for the soft and hard source cases. The numerical calculation and 1D PSTD transient simulations are conducted to verify the excellent performance of these optimum sources.