How to get the amplitudes along the MZO operator

The amplitude scheme that I use for the MZO operator is naive. It is roughly equal to Zhang's (1988) true amplitude DMO operator minus the phase correction. I plan to develop a fast algorithm to calculate the amplitudes along the MZO operator. The principle of the algorithm is based on the definition of the MZO operator, which is an operator that transforms constant-offset data to zero-offset data. Given a single diffractor in a velocity model, we can model the zero-offset diffraction, convolving the kinematic curve with a particular wavelet. We do the same for a constant-offset section and obtain the flat top pseudo-hyperbola. The MZO operator applied to the flat top pseudo-hyperbola should transform the section into a zero-offset section. We know the kinematics of the MZO operator, which implies we know the input samples from the constant-offset section which are summed together in an output point in the zero-offset section. For each point in the zero-offset section, we have a number of input values along a particular impulse response of the MZO operator. To determine the amplitude for the input points along the MZO operator, we have to solve a linear system of equations, or a least-squares problem by taking as many output values as necessary from the zero-offset section and the corresponding input values along a particular MZO operator in the constant-offset section. We don't have to solve the system of equations for each output time sample, as we can later interpolate for the full operator.