Saddle point method for transient processes in waveguides

A modification of the saddle point method is proposed for computation of non-stationary wave processes (pulses) in waveguides. The dispersion diagram of the waveguide is continued analytically. A set of possible saddle points on the dispersion diagram is introduced. A method of checking whether the particular saddle points contribute terms to the field decomposition is proposed. A classification of the waveguides based on the topology of the set of possible saddle points is outlined.