Optical DQPSK Modulation Performance Evaluation

Abstract

The always increasing amount of internet traffic in optical networks led to the need of using efficient modulation formats. Conventional on-off-keyed (OOK) signals have been extensively employed in optical communication systems. However, OOK modulation is inadequate for transmission of 40 Gbit/s per channel, or higher, bit rates (Winzer & Essiambre, 2006) mainly because of its reduced robustness to fiber nonlinearity, chromatic dispersion and optical filtering at such bit rates. In order to overcome such impairments, several advanced modulation formats received particular attention in the last few years. Some of these modulation formats still carry the information in the amplitude of the signal. However, they also modulate the phase of the signal to increase its robustness to transmission impairments. Some examples of such modulation formats are duobinary and alternate mark inversion. The most promising modulation formats for future optical networks make use of the phase of the signals to carry information. Among such formats, differential phase-shift-keying (DPSK) and differential quadrature phase-shift-keying (DQPSK) are the ones more often referred. These modulation formats led already to several landmark experimental results confirming their potential (Ho, 2005), (Winzer et al., 2008). The main advantages of DPSK are approximately 3 dB improvement on optical signal-to-noise ratio (OSNR) when compared with conventional OOK, and improved dispersion and polarization mode dispersion tolerance (Xu et al., 2004). DQPSK shows also improved spectral efficiency (Morita & Yoshikane, 2005). The evaluation of DQPSK system performance is usually performed using Monte-Carlo (MC) simulation or Karhunen-Loeve series expansion (Bosco & Poggiolini, 2006). The complexity of evaluating the performance of DQPSK system where noise added by optical amplifiers is the main noise source is the main reason for using these performance evaluation methods. However, while MC simulation has the time consumption disadvantage, Karhunen-Loeve series expansion derivation is usually of complex nature when rigorous DQPSK system performance evaluation is the goal. Furthermore, these methods provide reduced insight on the effects impairing the transmission system performance unless extensive analysis of different sets of parameters is performed.