Page 372 - Phase Space Optics Fundamentals and Applications
P. 372

Phase Space in Ultrafast Optics    353


                 An important feature of the paraxial approximation is that it is
               straightforward to propagate Gaussian beams using the transfer ma-
               trices acting on the complex beam parameter

                                      1    1   i
                                        =   −                      (11.38)
                                      q   R    w 2
               In Eq. (11.38), R is the radius of curvature of the beam at the reference
               plane, w is the corresponding beam size, and   is the index-dependent
               wavelength in the medium. The complex beam parameters before and
               after propagation, q and q , respectively, are linked by the formula

                                      q     Aq/n + B
                                        =                          (11.39)
                                      n    Cq/n + D
               where n and n are the optical index in the medium before and

               after propagation, respectively. The elements of the transfer matrix
               of Eq. (11.28) are then interpreted as modifying the beam waist and
               radius of curvature, accordingly.


               11.2.5 Temporal Paraxiality and the
                        Chronocyclic Phase Space
               An optical pulse may be represented in a manner similar to optical
               rays in geometrical optics. The analogy between space and time, the
               space-time duality, has been very fruitful. 36−38  Consider the action of
               a linear filter on a pulsed field. The relationship between input and
               output fields for the filter is





                                E out (t) =  dt H(t, t )E in (t )  (11.40)
               which can also be written in the frequency domain as

                                             ˜


                              ˜ E out ( ) =  d  ˜ H( , −  ) ˜ E in (  )  (11.41)
               Let us define a temporally paraxial approximation and postulate a
               general linear filter function in the form of a temporal Fresnel kernel

                                 1         i
                                               2          2
                       H(t, t ) = √  exp −   (at − 2tt + dt )      (11.42)
                                 2 b       2b
               where a, b, and d are real numbers. Here H is unitary and verifies



                                 dt H(t, t )H (t, t”) =  (t − t”)  (11.43)
                                           ∗
   367   368   369   370   371   372   373   374   375   376   377