Page 87 - Phase Space Optics Fundamentals and Applications
P. 87

68   Chapter Three


                 If a complex field amplitude f (r) is canonically transformed with
               the matrix T, then its Fourier spectrum is canonically transformed
                      t −1
               with (T )  . It is easy to see that for the case of phase-space rotators
               both transforms coincide.
                 For the description of phase-space rotations we can use the uni-
               tary matrix U = X + iY instead of the ray transformation matrix T O .
               Indeed, by introducing the following complex vector q = r − ip,it
               is easy to check that the ray transformation equation, expressed in
               dimensionless variables, Eq. (3.2), for the case of orthogonal matrix


                                   r o     X   Y   r i
                                       =                            (3.20)
                                   p o    −Y   X   p i
               can be rewritten as


                                (r − ip) o = (X + iY)(r − ip) i     (3.21)
               orinmorecompactformasq o = Uq i .Thispresentationunderlinesthe
               similarity in the description of phase-space rotators and polarization
               rotators of the monochromatic paraxial beams defined by correspond-
               ing Jones matrices. 10
                 Further, the RCT operator associated with unitary matrix U will
                             U
               be denoted as R . As well as for the ray transformation matrix T,
               the additivity of the phase-space rotators is expressed as R U 2  R U 1  =
                R U 2 ×U 1 .
                 Since matrix U has four free parameters, there are four uniparamet-
               ric groups of phase-space rotators: symmetric fractional FT, Eq. (3.15),
                 x =   y ; antisymmetric fractional FT, Eq. (3.15),   x =−  y ; gyrator,
               Eq. (3.16); and signal rotator, Eq. (3.14). These transforms are often
               written in the form of Hermitian operators 20,24−26
                                      !                "
                                     1  2   2    2    2
                                J 0 =
                                .      . x + . y + / p x + / p y
                                     4
                                      !                "
                                     1  2   2    2    2
                                J 1 =  . x − . y + / p x − / p y
                                .
                                     4
                                     1
                                J 2 =
                                .      . x. y + / p x / p y
                                     2
                                     1
                                J 3 =
                                .      . x/ p y − . y/ p x          (3.22)
                                     2
               where . x and . y, / p x =−i∂/∂x, and / p y =−i∂/∂y are position and mo-
               mentum operators. The operators J 0 , J 1 , J 2 , and J 3 are associated with
                                            . . .
                                                       .
               symmetric and antisymmetric fractional Fourier transforms, gyrator,
               and rotator, respectively. Note that the operator J 0 commutes with all
                                                       .
               others and that [J i ,J j ] = iε ijk J k , where i, j, k = 1, 2, 3 and ε ijk is the
                              . /
                                        .
               totally antisymmetric symbol, normalized through ε 123 = 1.
   82   83   84   85   86   87   88   89   90   91   92