The spectral gap of random regular graphs

From MaRDI portal



Abstract: We bound the second eigenvalue of random d-regular graphs, for a wide range of degrees d, using a novel approach based on Fourier analysis. Let Gn,d be a uniform random d-regular graph on n vertices, and let lambda(Gn,d) be its second largest eigenvalue by absolute value. For some constant c>0 and any degree d with log10nlldleqcn, we show that lambda(Gn,d)=(2+o(1))sqrtd(n−d)/n asymptotically almost surely. Combined with earlier results that cover the case of sparse random graphs, this fully determines the asymptotic value of lambda(Gn,d) for all dleqcn. To achieve this, we introduce new methods that use mechanisms from discrete Fourier analysis, and combine them with existing tools and estimates on d-regular random graphs - especially those of Liebenau and Wormald.











This page was built for publication: The spectral gap of random regular graphs

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6076727)