Analysis of Jitter Impact on High Speed Transmissions of Wavelength-Division Multiplexing Networks

: In this study, we conduct a thorough assessment of the effect of jitter occurrence in high speed 10 Gbps and 200 GHz Wavelength-Division Multiplexing (WDM) optical network. First, we present a simulation model to study the effect of jitter presence in the proposed network and then determine the maximum amount of jitter which the network can withstand. The model is then employed to predict the types of jitter received at the end of the transmission line. For the input power level of 0 dBm and Bit Error Rate (BER) of 1E − 09 , the observed total jitter, J T , random jitter, J R and deterministic jitter, J D is 0.2676 UI, 0.1602 UI and 0.1073 UI, respectively.


Introduction
Timing Jitter is one of serious issues in the voice and data transmission which directly contributes to signal degradation in high speed data communications systems. At transfer rates exceed 1 Gbps, digital signal appears to resemble the nature of analog signals. Jitter can be defined as the deviation of received bits as compared to a reference clock, thus leads to bit errors in signal transmission. It is known that, jitter appears to be one of important limitations to the performance of data transmission especially in optical Wavelength Division Multiplexing (WDM) system. Measurement of timing jitter or phase distortion is important to determine the sources, hence further precautions can be taken to reduce the jitter generation.
The overall jitter, J T is composed of both deterministic, J D and Gaussian (random), J R content. J R is generally generated from laser source's random noise that produces thermal noise and shot noise (Hancock, 2004), while J D is not intrinsic or random. J D has a specific cause and often periodic and narrowband. For instance, deterministic noise may be caused by imperfection of the transmitter circuit, poor circuit termination, etc. Ultimately, J D is correlated to Inter-Symbol Interference (ISI), duty cycle distortion and bandwidth limitation (Corrigan, 2000).
The presence of jitter can be measured intuitively by eye diagram analyzer. Eye diagram displays a composite view of each transition points over the Unit Interval (UI). Deploying eye diagram analyzer, the relative magnitudes of J D and J R can be revealed through immediate visual of jitter characteristic profile. Various mathematical estimation and simulation approaches can be used to predict the performance of a network system with jitter existence. Our work applied the OptiSystem software from Optiwave to study the behavior of jitter in the high speed optical transmission link especially for Wavelength Division Multiplexing (WDM) network. Figure 1 displays the proposed network setup to test jitter tolerance along the optical transmission line. The transmitter section consists of a Pseudo Random Bit Sequence (PRBS) generator, a Non Return Zero (NRZ) pulse generator, a Continuous Wave (CW) laser, Mach-Zehnder Modulator (MZM) and a coupler. PRBS generator is used as the data source that provides the basis for the formation of binary pattern. Bit sequences are fed into a PRBS generator using the NRZ modulation format and Mach-Zehnder is used as a modulator which transforms the electrical signals to optical signals.  The CW laser launched an optical power of 0dBm with a data rate of 10 Gbps, at 1546 nm wavelength. At the receiver side, the incoming signal was fed to the 200 GHz optical filter with Microring Resonator (MRR) filtering structure. The PIN photo diode serves as the optical signal detector and converts it into the electrical signals, while Bessel type Low Pass Filter (LPF) was used to filter the electrical signals and the signal is then fed to jitter generator. The noise generated at the receivers is set to be random and uncorrelated and the performance of the network with the presence of jitter was evaluated by referring to the Bit Error Rate (BER).

Jitter Tolerance Evaluations
The test is performed to test the resilience of the system, particularly with the use of MRR devices as optical filters. The simulation values of the MRR filter was summarized as in Table 1, obtained from previous research works (Mulyanti et al., 2014;Haroon et al., 2012;2013). Figure 2 portrays the results of the eye diagram observed with the existence of various jitter values. Horizontal axis eye opening was narrowed with the increase of jitter in the system. The phenomenon restricted the transmission of the data samples and further contributes to bit error. The presence of jitter in optical communication systems will cause noise margin to decrease and as the result, the systems are more vulnerable to transmission errors.
From Fig. 2, it can be clearly seen that, as the jitter presence closes the data eye, the BER is increasing. With the existence of 1 UI jitter as shown in Fig. 2(f), the eye opening was completely closed. Through this research, the maximum jitter tolerance that the network can withstand as for the BER <1×10 −9 is 0.639 UI. A research conducted by Intel discovered that the jitter tolerance for 10 Gbps digital system with the specifications recommended by the ITU-T G.825 is 0.5 UI, in which any optical transmission systems operated under this specific value, will function properly without any high impact degradation (Christensen, 2000). Therefore, it can be concluded that the point-to-point optical network as shown in Fig. 1 is working under the permitted jitter tolerance.

Jitter Measurement and Discussion
Details research on the network performance with the presence of jitter has been performed with the network architecture as shown in Fig. 3. The realized WDM direct scheme has eight channels and consists of three parts: A transmitter, an optical fiber and a receiver. Eight channels with 200 Ghz channel distributions were combined into a single transmission line by a multiplexer where each signal carry a data rate of 10 Gbps and complied the ITU-T G.692 frequency grid recommendations (Iyer, 2011).
The fiber optic was set to ITU-T G.652 standard includes the attenuation, group delay, group velocity dispersion and effective index of refraction. Other parameters considered as per the practical scenario of WDM system were listed in Table 2.
Through the eye diagram analysis, the three types of jitter as explained before were calculated and plotted as in Fig. 4. It can be clearly seen that the resulted jitter was correlated with the BER value. It can be concluded that, as the BER value getting better, the amount of jitter will be decreased. It is interesting to note that the observed value of J R remains almost unchanged despite the variation of BER.   In WDM optical transmission system, the J R occurrence is due by a laser source which produces thermal noise and shot noise, whereas J D resulted from the white noise produced by the components in optical communication systems (Yang and Lee, 2004). Since the presence of J R was due to random noise and has no specific boundaries, it is complicated to determine the contributing factors, nevertheless the J D was predictable and the source that lead to the jitter presence can be determined and eliminated to improve the overall system performance. J D can either be reduced with a good network composition layout, perfect line termination or reduce the line crosstalk (Zarr, 2012).

Conclusion
We have successfully investigate the impact of jitter in high speed 10 Gbps and 200 GHz Wavelength-Division Multiplexing (WDM) optical network using commercially available software. In particular, we have focused on an 8 channel WDM network consists of a transmitter, an optical fiber and a receiver. The observed total jitter, J T , random jitter, J R and deterministic jitter, J D is 0.2676 UI, 0.1602 UI and 0.1073 UI, respectively.