Page 160 - Fiber Optic Communications Fund
P. 160

Optical Modulators and Modulation Schemes                                          141


           duration of the signal pulse is 30 ps, the duty cycle is 30%. The duty cycle of a NRZ signal can be considered
           to be 100%. In the above definition, we have assumed rectangular pulses. For pulses of arbitrary shape, the
           duty cycle x can be defined as the ratio of the FWHM of a pulse to the bit interval T ,
                                                                             b
                                                    FWHM
                                                x =       .                                  (4.2)
                                                      T
                                                       b
            When rectangular pulses are used, a RZ pulse in a bit interval [−T ∕2, T ∕2] may be written as
                                                                 b
                                                                      b
                                            p(t)= 1for |t| < xT ∕2
                                                            b
                                               = 0 otherwise.                                (4.3)

           4.4  Power Spectral Density
           In this section, we find an expression for the power spectral density (PSD) of various line coders. Let the
           message signal be of the form
                                                      L
                                                     ∑
                                         m(t)= A lim     a p(t − lT ).                       (4.4)
                                                0         l     b
                                                 L→∞
                                                     l=−L
           Noting that
                                          {p(t − lT )} = ̃p(f)e i2flT b ,                 (4.5)
                                                   b
           the Fourier transform of m(t) is
                                                         L
                                                        ∑
                                         ̃ m(f)= A ̃p(f) lim  a e i2flT b .                (4.6)
                                                             l
                                                0
                                                    L→∞
                                                        l=−L
           The PSD is defined as
                                                             2
                                                      < | ̃m(f)| >
                                            (f) = lim         ,                            (4.7)
                                            m
                                                  T→∞     T
           where T =(2L + 1)T and <  > denotes the ensemble average. From Eq. (4.6), we have
                            b
                                    2
                                            ∗
                               | ̃m(f)| = ̃m(f) ̃m (f)
                                                  L              L
                                                 ∑              ∑
                                                             ∗
                                        2
                                     = A ̃p(f) lim  a e i2flT b  ̃ p (f)  a e
                                                                     ∗ −i2fkT b
                                        0            l               k
                                             L→∞
                                                l=−L            k=−L
                                                    L  L
                                                   ∑ ∑
                                        2
                                              2
                                     = A |̃p(f)| lim      a a e      .                       (4.8)
                                                            ∗ i2f(l−k)T b
                                        0      L→∞         l k
                                                   l=−L k=−L
           Using Eq. (4.8) in Eq. (4.7), we obtain
                                                         L  L
                                                  1     ∑ ∑         ∗
                                    2
                                         2
                             (f)= A |̃p(f)| lim              < a a > e i2f(l−k)T b .     (4.9)
                             m      0     L→∞ (2L + 1)T           l k
                                                      b l=−L k=−L
           Let us first consider the case of a polar signal in which a is a random variable that takes values ±1 with equal
                                                       l
           probability. When k ≠ l,
                                                     ∗
                                                < a a >= 0.                                 (4.10)
                                                   l k
                                            ∗
           This can be explained as follows. When a a = 1, it corresponds to a = 1 and a = 1, or a =−1 and a =−1;
                                                                k
                                                                                           l
                                                                         l
                                                                                 k
                                           l k
                  ∗
                                                                                        ∗
           when a a =−1, it corresponds to a =−1 and a = 1, or a = 1 and a =−1. The chance that a a = 1isthe
                                                         k
                                                 l
                                                                  l
                                       k
                l k                                                                   l k
   155   156   157   158   159   160   161   162   163   164   165