Open Access
Add to BibSonomyAbstract
An all-optical 2-to-4 level encoder based on cross polarization modulation in a semiconductor optical amplifier utilized to develop a 2 input digital multiplexer is presented. The most important feature of this device is the potential it has to improve the use of the available channel bandwidth on communications systems on an all-optical environment.
©2006 Optical Society of America
1. Introduction
In an optical digital communication system, the modulator generally maps an electrical sequence of binary digits into a set of corresponding optical signal waveforms. These waveforms may differ in either, amplitude, phase, frequency, polarization state or some combination of two or more signal parameters. When the mapping or assignment of k information bits is made to M=2k possible optical signal amplitudes, the modulation is called amplitude-shift keying (ASK). When k≥2, the modulation scheme is called multi-level amplitude-shift keying. Nowadays, most of the optical communications systems utilize amplitude-shift keying (ASK) modulation, nevertheless, the massive utilization of the Internet and multimedia applications demand a more efficient exploitation of the transmission channel bandwidth. This can be achieved by utilizing optical time or wavelength division-multiplexing (OTDM or WDM), as well as a multi-level amplitude-shift keying (MASK) modulation, which increases the bit rate without expanding the spectral width. Yet, a more efficient way to exploit the transmission bandwidth would be to implement MASK modulation on each channel of WDM systems. Similarly, a digital multiplexing of several channels in a WDM system can result in an enhanced utilization of the transmission channel bandwidth [1, 2].
On the other hand, recently, applications based on nonlinear polarization rotation and cross-polarization modulation (XPolM) effects [3] in SOAs have been on the spotlight. Some of them comprise the implementation of all-optical header processing systems [4], all-optical logic gates [5, 6], wavelength converters [7], all-optical switches [8], all-optical flip-flop memories [9], etc.
In this letter, an all-optical 2-to-4 level encoder based on XPolM effect in a semiconductor optical amplifier (SOA) is proposed. This device can be utilized in a WDM system as an all-optical 2-inputs digital multiplexer whose operation will be experimentally demonstrated.
2. Operation principle
In general, the XPolM effect changes the polarization-state of a CW beam as a function of the instantaneous optical power generated by one or several optical signals introduced simultaneously with it into a SOA. Based on that principle, an all-optical 2-to-4 level encoder will be carried out. In this case, the CW beam will be injected into the utilized SOA together with two synchronized digital optical signals of 2-level whose high-states produce different powers inside the amplifier active region, which will generate four possible optical powers. Therefore, four different polarization-states will be induced on the electric field of the CW beam at the SOA output through XPolM effect. Thus, if the CW beam leaving the SOA is introduced into a polarizer, it will be possible to generate at the output of this device, a signal with four different power levels, one for each polarization-state. However, it is important to notice, that strictly speaking, the signal being introduced into the polarizer is no more a CW beam since the two binary digital signals generating the four optical powers inside the SOA also change the gain of this device. Consequently, the power level of the CW beam is modified inside the SOA through cross-gain modulation (XGM) effect. Hence, for obtaining a single 4-level signal containing the information comprised in the two binary digital optical signals, it will be indispensable to take into account the action of the polarizer on the amplitude of the beam crossing it as well as the effect of the XGM on the CW beam in the SOA. That is to say, that for implementing an all-optical 2-to-4 level encoder, it will be necessary to find a kind of transfer function, or more precisely, an output power function for the CW beam, where the effects of the XPolM together with those of the polarizer and the XGM are considered.
In order to calculate the output power function, we propose the next procedure. First, for a defined optical power and a precise input polarization-state of the CW beam, we have to determine the evolution of its polarization-state at the SOA output as a function of the total optical power existing at the amplifier input. Next, utilizing this result we have to estimate the evolution of the CW beam amplitude after crossing a polarizer. To that aim, we propose to obtain, for a specific orientation of the polarizer, the transmission factor, affecting the input power of the beam crossing this device, as a function of the SOA input total power (i.e. as a function of the polarization-state of the CW beam at the SOA output). After that, we have to determine the evolution of the CW beam power at the SOA output (governed by the XGM) as a function of the total optical power at the amplifier input. Finally, we have to multiply this result by the transmission factor for each studied SOA input total power in order to obtain the output power function for the CW beam as a function of the SOA input power. Evidently, this output power function is valid only for a specific power and a determined polarization-state of the CW beam at the SOA input.
3. Device design
In order to find the output power function for the CW beam, that allows implementing the all-optical 2-to-4 level encoder, the XPolM and XGM effects were characterized utilizing the experimental setup shown in Fig. 1.
In this setup, it was utilized a single input signal with tunable power (signal 1) for emulating the total power generated by the addition of the two input digital signals that are introduced into the SOA for implementing the all-optical 2-to-4 level encoder. An erbium doped fiber amplifier (EDFA) is included for compensating the losses in the optical elements of the setup. The amplified signal 1 is polarization controlled and introduced into a traveling-wave SOA, whose structure is based on an InGaAsP/InP ridge waveguide of 1.5 mm, through an optical circulator. By the other side of the SOA, a CW beam with a power of 150 µW is polarization controlled and injected. The waveform of the CW beam, modified by the XPolM and XGM inside the SOA, is sent towards the circulator output. Next, a 0.9 nm band-pass filter is included in order to minimize the spontaneous emission of the SOA. At the filter output a polarization controller is placed followed by a 3 dB coupler whose output ports are connected to an optical spectrum analyzer and a polarization analyzer.
Fig. 3. Transmission factor evolution of the CW beam calculated at the output of an ideal polarizer vs. signal 1 power.
The aim of the experiment is first, to determine the evolutions of the power and the polarization-state of the CW beam at the SOA output (the former governed by the XGM and the second by the XPolM) as a function of the signal 1 power. Next, knowing the evolution of the polarization-state of the CW beam at the SOA output, to estimate the evolution of the transmission factor for this beam at the output of an ideal polarizer, represented in Fig. 1 as virtual polarizer. Finally, to determine the output power function for the CW beam, by multiplying the power evolution governed by the XGM (see Fig. 2) and the evolution of the transmission factor calculated at the ideal polarizer output (see Fig. 3). The result is shown in Fig. 4.
In this experiment, the signal 1 was introduced into the SOA with a wavelength of 1560 nm and a linear polarization-state oriented at 0 degrees (TE), while the CW beam was injected with a wavelength of 1554 nm and a linear polarization-state oriented at 45 degrees. The orientations of the polarization-states of these signals were chosen at 0 and 45 degrees for obtaining the best amplification of the signal 1 and the strongest effect of the induced birefringence on the CW beam, respectively. The power level of signal 1 was varied at the SOA input from 0 to 1.6 mW in 22 steps. Furthermore, the polarization-state of the CW beam at the filter output was converted to one horizontal linear when the signal 1 power was turned off and the transmission axis of the ideal linear polarizer was oriented vertically. As it is possible to verify on Fig. 4, there are four equidistant output powers of the CW beam (separated from each other by 22.5 µW) obtained for four different power levels of the signal 1 (P1, P2, P3 and P4), keeping the next relation: the value of the chosen highest power level for the signal 1 (P4) is equal to that resultant of the sum of the two selected intermediate power levels (P4=P2+P3). This issue is important, since in an actual all-optical 2-to-4 level encoder, the signal 1 will be substituted by two input signals (Q1 and Q2) whose amplitudes will be modulated digitally (in low and high levels). Thus, for obtaining an encoded signal of four power levels, it will be necessary that the high levels of the Q1 signal produce, inside the SOA active region, a different power from that produced by the high levels of the Q2 signal.
Fig.4. . W beam power at the ideal polarizer output vs. signal 1 power. This Fig. represents the power function, for the CW beam, calculated at the encoder output.
Fig. 5. Experimental setup utilized for demonstrating the operation of an all-optical 2 input digital multiplexer.
Particularly, for obtaining an encoded signal of four equidistant power levels, it will be necessary, according to Fig. 4, that the low and high power levels of the input digital signals are 0 and 0.5 mW for the Q1 signal and 0 and 0.76 mW for the Q2 signal. Evidently, this can be achieved by adjusting the power of the high levels of the input digital signals but also by tuning the wavelengths of these signals. In fact, the SOA gain is dependent of the wavelengths of the input digital signals. Consequently, the power generating the high levels of the input signals, inside the amplifier active region, depends on their wavelengths. Thus, it is possible to obtain an encoded signal of four equidistant power levels with input digital signals presenting the same power in their high levels.
On the other hand, it is interesting to notice, that according to Fig. 4, the low power level of the input digital signals can fluctuate between 0 and 0.06 mW for achieving an encoded signal of four equidistant power levels, which authorizes the utilization of input digital signals with moderate extinction ratios.
4. Experimental results
Figure 5 shows the experimental setup utilized for demonstrating the operation of an all-optical 2-to-4 level encoder utilized as an all-optical 2 input digital multiplexer. In the experiment, a 3 dB optical coupler and two tunable lasers were utilized along with two electro-absorption modulators for generating the two input digital signals. The wavelengths of these signals were fixed at 1560 and 1558 nm. Utilizing a pattern generator, one of the signals (Q1) was modulated with a periodic rate of 5 Gb/s (in NRZ code) and the other (Q2) was modulated with a 2.5 Gb/s NRZ code with a pseudo-random data sequence of 223 - 1 length. These binary signals were amplified via an erbium doped fiber amplifier (EDFA) and their high and low levels were fixed at 0.05 mW and 0.55 mW respectively. The polarization-states at the SOA input of these signals were converted to one linear oriented at 0 degrees. For these conditions, the gain peak of the SOA was in 1557.5, which allowed that the high levels of the Q2 input digital signal, (whose wavelength is 1558 nm), were more amplified than those of the Q1 input signal. Thus, it was possible to generate an output signal of four levels since the high levels of the Q1 signal produce, inside the SOA active region, a different power from that produced by the high levels of the Q2 signal.
In the experiment, a DFB laser emitting with a wavelength of 1554 nm was used to provide the CW beam. The power of this beam at the SOA input was of 150 µW and its polarization-state was converted to one linear oriented at 45 degrees. After passing by the SOA, the CW beam was introduced into a circulator and sent towards a band-pass filter. Next, utilizing the polarization controller PC3, the polarization-state of this beam was converted to one horizontal linear when the Q1 and Q2 signals presented simultaneously, at the SOA input, a low level. The PC3 output was sent towards a Glan-Thompson polarizer, with an extinction ratio of 100,000:1, whose transmission axis was oriented vertically and the output of this device was received on a photodetector, which was connected to a sampling scope.
Figure 6 shows the experimental demonstration of the all-optical 2-to-4 level encoder utilized as an all-optical 2 input digital multiplexer. In this Fig. it is possible to confirm that the device produces an output signal of four equidistant power levels when at its input are introduced two digital signals modulated with high and low levels of similar power. In this Fig., the lowest power level of the output signal (P1) is obtained when both input digital signals present a low level (0, 0). The intermediate power levels of the output signal (P2 and P3) are obtained when the input digital signal modulated with a periodic rate of 5 Gb/s presents a digital state different to that presented by the signal modulated with a 2.5 Gb/s pseudo-random data sequence. In this case, when the first signal is in a low level, the other is in a high level (0,1) and the output signal presents a P2 power level. Inversely, when the first signal is in a high level, the other is in a low level (1,0) and the output signal presents a P3 power level. Finally, the highest power level of the output signal (P4) is obtained when both input digital signals present a high level (1, 1).
Since, the encoded output signal contains in four power levels the information being presented at the same time by the two input binary signals, the Fig. 6 also represents the experimental demonstration of an all-optical 2 input digital multiplexer.
It is important to notice that for the above described device to operate correctly, it is necessary that the data rates of the two input digital signals be equal, and the rise and fall edges of their bits arrive at the same time (in phase) at the SOA input. Otherwise, the minimum duration (which is equal to the bit period of the fastest input signal) of the output levels will be variable and eventually some of the levels will be lost. In fact, the device is very sensible to a weak constant difference in the data rates of the two input digital signals, since it provokes a permanent sliding of the bits, which desynchronizes the system and produces output levels whose minimum duration will fluctuate in the time. Nevertheless, if the weak difference in the data rates of the two input digital signals is random in relation to a central data rate, as it is usually the case, the sliding of the bits will be both positive and negative and its effect will be compensated in the course of time. Therefore, the minimum duration of the output levels will be kept at an average value. On the other hand, if the rise and fall edges of the bits of the two input digital signals arrive at the SOA input with a weak phase difference (for instance, smaller than an eighth of the bit period of the fastest input signal), output levels of small duration that will not contain valid information will be generated; however it will be always possible to obtain the right information by sampling at an adequate time. That is to say, by sampling with a delay, (in relation to any rise edge preceding to any output level), whose value is equal to half the bit period of the fastest input signal.
Fig. 6. Experimental demonstration of the all-optical 2-to-4 level encoder based on cross polarization modulation utilized as a 2 input digital multiplexer. P1, P2, P3 and P4 represent the four power levels of the output signal of the digital multiplexer.
5. Conclusion
An all-optical 2-to-4 level encoder based on cross polarization modulation in a semiconductor optical amplifier utilized to develop a 2 input digital multiplexer was experimentally demonstrated. The experiment consisted in digitally multiplexing two binary waveforms, of 1560 and 1558 nm, in a single four-level signal of 1554 nm. The experimental demonstration was realized with two digital signals modulated with NRZ code. The former, with a periodic rate of 5 Gb/s and the second with a 2.5 Gb/s pseudo-random data sequence of 223 - 1 length.
The results obtained in this work have demonstrated that this device can be utilized by improving the use of the available channel bandwidth on communications systems on an all-optical environment.
Acknowledgments
This work was partially financed by the Consejo Nacional de Ciencia y Tecnología de México (CONACYT).
References
1. L. Huo, J. Zhao, C. Lin, C-K. Chan, and Z. Wang, “Experimental demonstration of a novel all-optical multilevel 4-amplitude-shifted-keying coding/decoding scheme,” OFC/NFOEC 2006, paper JThB41 (2006).
2. N. Kikuchi, “Amplitude and phase modulated 8-ary and 16-ary multilevel signaling technologies for high-speed optical fiber communication,” SPIE-OMDOC 6021, 568–575 (2005).
3. H. Soto, D. Erasme, and G. Guekos, “Cross-polarization modulation in semiconductor optical amplifiers,” IEEE Photon. Technol. Lett. 11, 970–972 (1999). [CrossRef]
4. N. Calabretta, Y. Liu, F.M. Huijskens, M.T. Hill, H. de Waardt, G.D. Khoe, and H.J.S. Dorren, “Optical signal processing based on self-induced polarization rotation in a semiconductor optical amplifier,” J. Lightwave Technol. 22, 372–381 (2004). [CrossRef]
5. H. Soto, E. Alvarez, C.A. Díaz, J. Topomondzo, D. Erasme, L. Schares, L. Occhi, G. Guekos, and M. Castro, “Design of an all-optical NOT XOR gate based on cross-polarization modulation in a semiconductor optical amplifier,” Optics Commun. 237, 121–131 (2004). [CrossRef]
6. L.Q. Guo and M.J. Connelly, “All-optical AND gate with improved extinction ratio using signal induced nonlinearities in a bulk semiconductor optical amplifier,” Opt. Express 14, 2938–2943 (2006). [CrossRef] [PubMed]
7. L.Q. Guo and M.J. Connelly, “Signal-induced birefringence and dichroism in a tensile-strained bulk semiconductor optical amplifier and its application to wavelength conversion,” J. Lightwave Technol. 23, 4037–4045 (2005). [CrossRef]
8. H. Ju, S. Zhan, D. Lenstra, H. de Waardt, E. Tangdiongga, G.D. Khoe, and H.J.S. Dorren, “SOA-based all-optical switch with subpicosecond full recovery,” Opt. Express 13, 942–947 (2005). [CrossRef] [PubMed]
9. H.J.S. Dorren, D. Lenstra, Y. Liu, M.T. Hill, and G.D. Khoe, “Nonlinear polarization rotation in semiconductor optical amplifiers: Theory and application to all-optical flip-flop memories,” IEEE J. Quantum Electron. 39, 141–148 (2003). [CrossRef]
