A convex conic underestimate of Laplacian spectra and its application to network synthesis
Jan Maximilian Montenbruck, Alexander Birk, Frank Allgöwer · 2015
We derive sufficient conditions on the eigenvectors and eigenvalues of a matrix for letting it be the Laplacian of an undirected, weighted graph. The derived conditions are convex conic and apply to network synthesis problems. We present an algorithmic implementation of the proposed synthesis procedure.