A stochastic differential equation driven by N-order nonlinear fluctuation is investigated by defining a novel stochastic process Γ(t). The spectral densities of Γ(t) for N = 1, 2, 3 and 4 are obtained. We compare the results of the linear fluctuation case with the nonlinear fluctuation case and find that the extremum of the spectral density for linear case is directly proportional to the noise intensity D and independent of the correlation time τ, the extremum of the spectral density for nonlinear case is determined by both D and τ. An app...
