Page 284 - Fiber Optic Communications Fund
P. 284

Optical Amplifiers                                                                 265


                     √
           where |r | =  R , j = 1, 2. After j round trips, the partial field at B is
                        j
                 j
                                                      √           2j+1
                                                    j
                                       =  t t (r r ) [ G exp(i )]  .                 (6.115)
                                        j   in 1 2  1 2  s     0
           The total field output at B is the superposition of partial fields,
                                           ∞                       ∞
                                          ∑            √          ∑   j
                                      =     = t t   G exp(i )  h ,                 (6.116)
                                      out     j   1 2 in  s     0
                                          j=0                     j=0
           where
                                             h = r r G exp(i2 ).                         (6.117)
                                                 1 2
                                                     s
                                                             0
           The summation in Eq. (6.116) is a geometric series and if
                                                   |h| < 1,                                (6.118)
           we have
                                                ∞
                                                ∑   j   1
                                                   h =     .                               (6.119)
                                                       1 − h
                                                j=0
           Therefore, Eq. (6.116) becomes
                                                     √
                                                 t t  G s
                                                 in 1 2
                                           out  =       exp(i ).                       (6.120)
                                                               0
                                                   1 − h
           The overall gain G is defined as
                                                 2        2  2
                                                        1
                                              out
                                                            2
                                            | |      |t | |t | G (f)
                                                               s
                                      G(f)=       =                  .                     (6.121)
                                                2                ∗
                                             | |  [1 − h(f)][1 − h (f)]
                                              in
           Using Eq. (6.117 ), Eq. (6.121) can be rewritten as
                                                (1 − R )(1 − R )G (f)
                                                     1
                                                              s
                                                           2
                                  G(f)=                √                 .                 (6.122)
                                                 2
                                        1 + R R G (f)− 2 R R G (f) cos (2 )
                                            1 2  s        1 2  s        0
           Using the relation
                                                             2
                                           cos (2 )= 1 − 2sin  ,                       (6.123)
                                                 0             0
           Eq. (6.122) can be put in a different form:
                                                (1 − R )(1 − R )G (f)
                                                     1
                                                              s
                                                           2
                                     G(f)=                            ,                    (6.124)
                                                             2
                                                   2
                                           (1 − RG ) + 4RG sin (2nfL∕c)
                                                 s       s
                    √
           where R =  R R is the geometric mean of facet reflectivities. From Eq. (6.124), we see that the peak gain
                       1 2
           occurs when
                                        2nfL
                                              = m,  m = 0, ±1, ±2, …                     (6.125)
                                          c
           or
                                                      mc
                                                  f =    ,                                 (6.126)
                                                  m
                                                      2nL
           which is the same as the resonant frequency given by Eq. (3.44). Therefore, the cavity-type optical amplifier
           amplifies any input signal whose frequency is matched to the resonant frequency f of the cavity. The peak
                                                                             m
   279   280   281   282   283   284   285   286   287   288   289