Page 42 - Phase Space Optics Fundamentals and Applications
P. 42

Wigner Distribution in Optics   23


               1.6.5 Transport Equations
               With the tools of this section, we could study the propagation of the
               Wigner distribution through free space by considering a section of
               free space as an optical system with an input plane and an output
               plane. It is possible, however, to find the propagation of the Wigner
               distribution through free space directly from the differential equation
               that the signal must satisfy. We therefore let the longitudinal variable
               z enter into the formulas and remark that the propagation of coherent
               light in free space (at least in the Fresnel approximation) is governed
               by the differential equation (see, for instance, Ref. 15, p. 358)
                                                  2
                                    ∂ f        1 ∂
                                  −i   =   k +        f             (1.55)
                                    ∂z        2k ∂r 2
                                                                2
                                                     2
                                                         2
                        2
                                                            2
                     2
               with ∂ /∂r representing the scalar operator ∂ /∂x +∂ /∂y and with
               k the wave number. The propagation of the Wigner distribution is now
               described by a so-called transport equation 7,8,67–70  which in this case
               takes the form
                                       t
                                    2 q ∂W    ∂W
                                            +     = 0               (1.56)
                                     k   ∂r    ∂z
               with ∂/∂r =∇. The transport equation (1.56) has the solution

                                                2 q
                              W(r, q; z) = W r −   z, q;0           (1.57)
                                                 k
               which is equivalent to the result Eq. (1.44) in Sec. 1.6.1, with the special
               choice A = D = I.
                 In a weakly inhomogeneous medium, the optical signal must satisfy
               the Helmholtz equation

                                    ∂ f            ∂ 2
                                           2
                                 −i    =   k (r,z) +  f             (1.58)
                                    ∂z             ∂r 2
               with k = k(r,z). In this case, we can again derive a transport equation
               for the Wigner distribution; the exact transport equation is rather com-
               plicated, but in the geometric-optical approximation, i.e., restricting
               ourselves to first-order derivatives, it takes the simple form

                             
                         	 t
                      t
                                2
                                       2 t
                   2 q ∂W      k − (2 ) q q ∂W      ∂k    ∂W
                           +                   +              = 0   (1.59)
                     k  ∂r          k       ∂z     2  ∂r   ∂q
               which, in general, cannot be solved explicitly. With the method of
               characteristics, however, we conclude that along a path defined by

                                               2 t
                                        2
                    dr  2 q     dz     k − (2 ) q q    dq    ∂k
                      =            =                      =         (1.60)
                   ds     k     ds         k           ds   2  ∂r
   37   38   39   40   41   42   43   44   45   46   47