Open Access
Add to BibSonomyAbstract
We propose and demonstrate a novel scheme for short pulse controlled all-optical switch using external cavity based single mode Fabry- Pérot laser diode (SMFP-LD). The proposed scheme consists of control unit and switching unit as two essential blocks. The basic principle of the proposed scheme is the optical bistability property of SMFP-LD for the control unit and the suppression of the dominant beam of SMFP-LD with injection locking for the switching unit. We also present the analysis of hysteresis width and rising/falling time with change in wavelength detuning which helps to find the optimum wavelength detuning value and power of light beams at different stages of the proposed scheme that gives wide input dynamic power range, high ON/OFF contrast ratio, and low rising/falling time. Input data of 10 Gb/s Non Return to Zero (NRZ) signal is switched at output ports depending upon the control signal generated by the control unit, which comprises of optical SR latch. Output waveforms, clear eye diagrams with extinction ratio of about 11 dB, rising/falling time of about 30 ps and 40 ps, and bit error rate (BER) are measured to validate proposed scheme. No noise floor is observed at output ports up to BER of 10−12 and the maximum power penalty recorded is about 1.7 dB at a BER of 10−9 which shows good performance of the proposed short pulse controlled optical switch using SMFP-LDs.
© 2014 Optical Society of America
1. Introduction
Optical signal processing and controlling has received considerable attention since it addresses various challenges such as information acquisition, transmission, storage and processing, and meets the demands of high speed and large capacity requirement of optical communication system. Additionally, optical signal possesses benefit of comparatively less electro-magnetic interference, less cross talk, no heat dissipation, no short circuit problems and many others [1, 2]. The technology advancement of high quality optical fiber has enabled the development of communication networks in present condition. It has not only improved communication but also has played a vital role for the development of related technology and other optical signal processing and controlling schemes. Optical computing which plays a significant role on the communication networks has been also developed as an integral part of it. Compared to the present electronics counterparts, photonics still need focus and extensive research and development on the improvement of efficient integration technology, power consumption, all-optical memory system and cost. Despite the present limitations, all-optical technology shows promising applications in the field of fast signal processing, higher accuracy sensing, control system in hazardous situations, and electromagnetic interference free system. One of these scenario is represented by optical interconnection between chip to chip used for short range photonics interconnection to improve the high performance computing [3]. The successful implementation of the optical interconnection in reality as in the PERCS (Productive, Easy-to-use, reliable Computing system) [4] and OSMOSIS (Optical shared memory supercomputer interconnect system) [5] have shown the feasibility and necessity of optical network. Optical bistability, an essential property for the realization of optical memory, have been analyzed by different research groups in many different approaches for the realization of optical buffers and RAM [6–10].
Optical switch is one of the important building block for the optical signal processing and communication system. Various techniques and components such as Mach-Zenhder interferometer, erbium doped fiber with long-period fiber gratings, semiconductor optical amplifier (SOA) have been used for the demonstration of optical switch and overcoming the limitation of each other [11–14]. Among various components, SOAs have shown promising developments on designing various optical units including switch. SOAs require high biasing current and are expensive [15]. To overcome the higher cost and high driving current requirements, optical units such as gates and switches have been proposed using commercially available multi-mode Fabry-Pérot laser diodes (MMFP-LD) [16–18] and especially designed single mode Fabry-Pérot laser diodes [19–21]. But all these techniques require input control signal to be high for the entire period of data to be blocked at the output ports.
In this paper, we propose and experimentally demonstrate an idea of controlling the switch using SMFP-LDs with short pulses input as a control signal. Only short pulses at the input control blocks are enough to switch and block input data at output ports. The switching window can be varied easily by implementing different periods of short pulses at the input of control signal. The control block is composed of optical SR unit which works on the bistability of SMFP-LD and the switching unit works on the principle of the injection locking with suppression of the dominant mode of the SMFP-LD. The hysteresis width analysis in injection locking of SMFP-LD with different wavelength detuning is done to find the optimum input power for the different stage of proposed switch which helps to provide the high dynamic range for the control unit and power efficient solution. Additionally, the effect of wavelength detuning on rising/falling time is also analyzed to provide minimum rising/falling time at the output ports. The proposed idea is verified experimentally by using 10 Gb/s data to be switched at output ports depending upon the short pulse applied on the control unit. The different duration of short pulse are implemented to show the performance of the proposed short pulse control switch. The 3.5 Gb/s NRZ data are used to provide the different duration of control pulse and to analyze the output performance with different pulse width of input control signal. Clear output waveforms, eye diagrams with extinction ratio of more than 11 dB, rising/falling time of about 30/40 ps and no noise floor up to BER of 10−12 are obtained at output ports.
2. Operating principle
The proposed scheme of short pulse controlled optical switch consist of two units, control unit and switching unit, as shown in Fig. 1. The control and switching unit work on optical bistability, injection locking, and gain modulation of SMFP-LDs [22–24]. In [23], optical logic units and comparator circuit using SMFP-LDs have been demonstrated but do not provide any insight and analysis about short pulse switching technique, effect of wavelength detuning on hysteresis width, and rising/falling time. In this paper, we analyze the effect of wavelength detuning on optical bistability and rising/falling time, and implement the results in demonstrating short pulse controlled switch using SMFP-LDs, which does not require input signals to be high for the entire duration of data transfer. SMFP-LDs, which are used in this scheme are specially designed and developed in our laboratory, have a dominant self-locked single longitudinal mode with high side mode suppression ratio [25]. The SMFP- LD is obtained by eliminating the inclinations of 6° to 8° of the coupling fiber present in conventional FP-LDs, thereby, forming an external cavity between laser diode and the fiber. The SMFP-LD consists of a FP-LD chip with a multi-quantum well of 300 µm and an external cavity length of 4 mm. By varying the temperature, the mode matching condition is achieved for both the cavities. The refractive index of the active region changes with the change in temperature. As a result, there is a change in optical path length in the laser diode providing the optimal mode matching condition for the single mode oscillation. This single mode oscillation can be tuned to another mode by varying the operating temperature, which gives the wavelength tunability of the dominant mode of the SMFP-LD. The required temperature stability is maintained with the commercially available temperature electric cooler (TEC) which has a temperature stability of 0.01°C. The self-locking mode of SMFP-LD is tunable to a wide range of wavelength of about 10 nm by varying the operating temperature. SMFP-LD shows similar characteristics as that of the MMFP-LDs with the injection of external beam. The difference between SMFP-LDs and MMFP-LDs is that the former does not need any external probe beam for signal processing.
Control unit has two inputs which are used to generate the control signal for switching input data at outputs based on the input control pulse applied. The control unit is composed of SR latch which gives the output based on the set and reset signal at the input. The output of SR latch, Q, which is the control signal for switching and blocking the input data at the output port 1 of switching unit, will be high with the implementation of set beam at the input of control unit. Output Q will maintain the same logic output, high, even when the set beam is removed. Output at Q will be logic 0 only when another control input beam, reset, is high with set as low. Hence the input data applied to switching unit will be switched to the output port 1 when Q is logic 0. Similarly the input data will be switched to the output port 2 when is logic 0.
Control unit comprises of two SMFP-LDs in the cascaded form where the output of SMFP-LD1 and set pulse are inputted to SMFP-LD2. It is noted that the set pulse is at the same wavelength as that of the self-injected mode of SMFP-LD1, λ1. This configuration helps to set the output of control unit i.e. Q as “1” even with lesser amount of power at set beam, which makes the configuration more efficient in terms of power requirement. The control output is obtained through SMFP-LD2. The output is divided into two and passed through different band pass filters, BPFs tuned at the wavelength of λ1 and λ2 which give opposite logic state of each other. These two beams Q and act as control signals for the corresponding output ports of the proposed switch with SMFP-LD3 and SMFP-LD4. Since control signals to switching unit are always in opposite logic state, the input data signal is either outputted to port 1 or port 2 based on the low/high state of Q and .
Figure 2 shows the spectrum schematic of the proposed short pulse controlled optical switch. Figure 2(a)-2(i) shows the spectrum schematic of self-injected SMFP-LD2 with dominant mode at the wavelength of λ2. At the initial condition both set and reset beams are logic “0”. At the initial condition, SMFP-LD2 has input beam of λ1 which is the dominant mode of the SMFP- LD1. The power of λ1 is managed in such a way that it is enough to injection lock the side mode of SMFP-LD2 but it is not sufficient enough to suppress the dominant mode of SMFP-LD2, λ2 as shown in Fig 2(a)-2(ii). The dominant mode of SMFP-LD2 will be suppressed only when there is a presence of additional beam which can be obtained from the set signal and is injected in the same wavelength as that of the dominant mode of SMFP-LD1 (λ1 = λs). As set beam uses the same wavelength as that of dominant mode of SMFP-LD1, small amount of power on the set pulse λs is sufficient enough to suppress the dominant mode of SMFP-LD2 as shown in Fig. 2(a)-(iii). It is noted that in order to suppress the dominant mode of SMFP-LD2, both beams; dominant mode of SMFP-LD1 and set beam are required. With previous state of suppression of the dominant mode of SMFP-LD2, the self-injected dominant beam of SMFP-LD1 alone is enough to hold on to the same state of λ2 as shown in Fig. 2(a)-2(iv). The suppressed state of λ2 will be in the same state until the reset beam is set to logic “1”. This gives the set sate of SR latch by providing Q which is in the wavelength of λ1 as logic high and which is in the wavelength of λ2 as logic low. Two outputs from SMFP-LD2 with Q and work as a control signal for the switching unit. With reset signal as logic “1”, the dominant mode of SMFP-LD1 will be suppressed and hence the input to the SMFP-LD2, which is after the BPF tuned at λ1, is logic “0” as shown in Fig. 2(a)-2(v). Hence the output from SMFP-LD2 will be logic “0” at Q and logic “1” at . In this case also, even with the removal of reset beam the output will maintain the same state since λ2 cannot be suppressed only with the presence of the λ1 which is shown in Fig. 2(a)-2(vi).
Fig. 2 Spectrum schematic of (a) control unit and (b) switching unit, of the proposed short pulse-controlled switch.
Two output state of Q and depend on the input control signals S and R which is used to pass the input signal either through port 1 or port 2. Since the configuration of the control unit for switching is SR latch the input control signals S and R are not required to be continuously in logic “1” state for transferring or blocking the data through output ports. Only a short pulse at the input S and R is sufficient enough for the operation of the switching unit. The switching unit works on injection locking and gain modulation of SMFP-LD which is illustrated in Fig. 2(b). Figure 2(b)-2(i) and 2(b)-2(ii) show the spectrum of dominant mode of SMFP-LD3 and injection locked SMFP-LD by data signal λdata which suppresses the dominant mode of SMFP-LD3, respectively. Furthermore, the gain or power of the injected data signal can be modulated by injecting another beam known as pump beam which is either Q or in the proposed scheme. When the control signal Q or is present the gain of the data signal will be suppressed based on the gain modulation principle. The amount of gain suppression depends on the power of the control signal which is directly proportional to the wavelength detuning of control signal to the corresponding mode of the SMFP-LD.
In order to find the optimum power and wavelength detuning for different beams of the proposed scheme, we analyzed the hysteresis width of the optical bistability, the suppression of the dominant mode of SMFP-LD, and rising/falling time of output signal which are shown in Fig. 3. The effect of wavelength detuning on the stability of injection locking with a single injected mode has been analyzed in [21] and it is shown that the negative wavelength detuning will result in unstable locking regime. However, with more than single injected locked modes, the negative wavelength detuning has relatively small region of instability but the negative wavelength detuning cannot provide bistability [26]. Hence, we have only considered the effect of positive wavelength detuning on hysteresis width in optical bistability and rising/falling time of output ports. The effect of positive wavelength detuning on optical bistabilty shown in Fig. 3(a) is matched with the theoretical analysis presented on [26]. One of the notable things that we have observed is the change in the width of the hysteresis. Figure 3(a)-3(i) and 3(a)-(3ii) shows the hysteresis curve with wavelength detuning of 0.02 nm and 0.04 nm. We can see that the hysteresis loop width is small in both cases and the power of the dominant mode drops sharply compared to that of Fig. 3(a)-3(ii) and 3(a)-3(iv) which have higher wavelength detuning. In addition, some noticeable issues are counterclockwise hysteresis with detuning of 0.02 nm, and a butterfly hysteresis loop with the increase in wavelength detuning to 0.04 nm. In Fig. 3(a)-3(iii), the wavelength detuning is maintained at 0.08 nm. When the injected power is set initially to a value of −6 dBm (marker1), the power of the dominant mode is about −5 dBm level. However, as the injected power is increased to −5.1 dBm, the dominant mode is fully suppressed as SMFP- LD is locked at the injected wavelength. On the contrary, decreasing the injected beam power below −5.1 dBm after the injection locking with suppression of dominant mode, the output of dominant mode maintain low output state (state of marker 2 and 3). The dominant mode will remain suppressed until the input injected beam is less than −6.2 dBm. On slightly decreasing the injected power from −6.2 dBm (marker 4), the dominant mode returns back to the unsuppressed mode and back to initial state (marker 1 again) gaining the power at the dominant mode and completes the hysteresis loop. It is noticeable that the corresponding modulation depth as a value of high-output to low-output states can reach a high level of 40 dB from Fig. 3(a)-3(iii). In addition, when the wavelength detuning is increased to 0.12 nm, the hysteresis loop width is further extended, which is illustrated in Fig. 3(a)-3(iv). It is observed that the optical bistability shown in Fig. 3(a) exhibits a transition from counterclockwise to clockwise hysteresis by the way of a butterfly hysteresis when the wavelength detuning is increased from 0.02 nm to 0.12 nm.
Fig. 3 Wavelength detuning effect on (a) hysteresis width (i) 0.02 nm (ii) 0.04 nm (iii) 0.08 nm and (iv) 0.12 nm, and (b) rising/falling time of output.
The effect of change in wavelength detuning on rising/falling time of the output port is illustrated in Fig. 3(b). The NRZ input signal is used for the analysis of the rising/falling time in SMFP-LD. Wavelength detuning is varied from the range of 0 nm to 0.14 nm. It is observed that the rising and falling time both increases with the increase in the wavelength detuning and the falling time is higher than rising time in SMFP-LDs. From Fig. 3(a), we observed that the increase in wavelength detuning increases the hysteresis width which in turn increases the rising and falling time of output of the SMFP- LD. With this result we can conclude that the higher wavelength detuning is preferred for the latch operation compared to the switching operation as wavelength detuning shows the directly proportional relationship with rising/falling time. Hence in our proposed approach, we used the higher wavelength detuning of 0.12 nm for the control unit which comprises of SR latch and lower wavelength detuning of 0.04 nm to the switching unit.
3. Experimental setup and results
The experimental set up for the proposed short pulse controlled all-optical switch is shown in Fig. 4. SMFP-LD1 and SMFP-LD2 generate control signals which are used to control the switching of input data from port 1 and port 2 via SMFP-LD3 and SMFP-LD4, respectively. SMFP-LD1 is cascaded to SMFP-LD2 in order to make it work as a SR flip-flop, which provides control signal for switching ports; Port 1 and Port 2. SMFP-LD3 and SMFP-LD4 work on the injection locking property and are used as a switching unit to switch the input data to output ports, according to the control signal. The logic state of control signal is determined based on the state of S, R and the previous control output state. Tunable lasers (TL), TL1 and TL2 are used to modulate short pulses which are control signals for switching input data to output port 1 or output port 2. Pulse pattern generators (PPG), PPG1 and PPG2 are used to generate 16 bit 3.5 Gb/s NRZ data and are optically modulated by the modulator, MOD, for controlling switching of input data to output ports. Data with different bit sequence are used to show the switching of input data signals at the output ports for different time interval with different input pulse width. The minimum width of pulse used for set and reset bit is about 250 ps. With S and R beam present for 250 ps, the proposed scheme can perform the switching operation to switch the input data to either output port 1 or output port 2 and maintain the data output to the same port even when the short pulse applied to control signal is absent until another control signal is applied. It shows the distinct difference between proposed short pulses controlled switch which require the control signal to be ON only for a short duration compared to the conventional optical switch which requires the control signal to be ON for the entire period of the switching operation.
Fig. 4 Experimental setup of the proposed short pulse-controlled switch using self-injected single mode Fabry-Pérot laser diode.
TL3 is used to modulate 10 Gb/s NRZ input data signal generated by PPG3 which is to be switched at the output ports, either to port 1 or port 2 based on the control pulse S and R. Polarization controllers (PC), PC1, PC2 and PC7 are used to minimize the loss in the Mach-Zehnder modulator where as other PCs are used to obtain the TE polarized light only to be injected on respected SMFP-LDs since SMFP-LDs works on the injection locking principle which is based on the TE polarized light. With TM polarized light source, absorption modulation phenomenon occurs [27]. SMFP-LD1, SMFP-LD2, SMFP-LD3, and SMFP-LD4 have threshold current of 9 mA, 11 mA, 13 mA, and 11 mA, and are biased with the biasing current of 14 mA, 17 mA, 20 mA, and 20 mA under the operating temperature of 21°C, 23.5°C, 26.7°C, and 17.5°C, respectively. Under these operating conditions, SMFP-LD1 to SMFP-LD4 have the dominant self-injection locked mode at the wavelength of 1546.2 nm, 1541.24 nm, 1544.2 nm, and 1539.84 nm with the side mode suppression ratio (SMSR) of 38.2 dB, 29.7 dB, 35.14 dB and 38.61 dB, respectively. BPF1 is used to pass the dominant mode of the SMFP-LD1, which is injected to SMFP-LD2. BPF2 and BPF3 are used to pass the wavelength of λ1 and λ2, which are control signals for SMFP-LD3 and SMFP-LD4. BPF4 and BPF5 are used at the output ports; Port1 and Port 2, which ensure only the wavelength of data signals are present on output ports.
Two control signals S and R are injected at the wavelength of 1546.2 and 1549.82 nm with the wavelength detuning of 0.12 nm with the respective SMFP- LD1 and SMFP-LD2. The 0.12 nm wavelength detuning is used for the control signal since from the analysis of Fig. 3(a), we found higher wavelength detuning gives wider hysteresis width which is desired for the SR latch operation. Set beam and dominant mode of SMFP-LD1 have the same wavelength so that even the set signal power of about −16 dBm with the presence of dominant mode of SMFP-LD1 (when R = 0) is sufficient enough to suppress the dominant mode of SMFP-LD2 for the Set state. This low power of set signal is possible due to the larger hysteresis width, since there is a large difference between the input power required for injection locking and releasing. Depending upon S and R pulse applied at the input of control unit, the modulated data which are inputted to both switching wings are either outputted to port 2 or port 1.
Figure 5(a) shows the spectrum diagram of the control unit that generates signals which are used for switching the inputted data to output ports. Figure 5(a)-5(i) and Fig. 5(a)-5(ii) show the spectrum output of SMFP-LD1 and SMFP-LD2 under the normal biasing condition of SMFP-LDs without any input beam injected to both SMFP-LDs. Figure 5(a)-5(iii) shows spectrum output of SMFP-LD2 without any input pulse S and R applied. But in this case, there is a presence of the dominant mode of SMFP-LD1 which is injected to SMFP-LD2. The power of the dominant mode of SMFP-LD1 is managed in such a way that the power is sufficient enough to weakly injection lock SMFP-LD2 but is not enough to suppress the dominant mode of SMFP-LD2. In order to suppress it, both inputs to SMFP-LD2, set signal, S, and output from SMFP-LD1, which is the dominant mode of SMFP-LD1, should be logic high. SMFP- LD2 is injection locked at the wavelength of λ1, 1546.86 nm as shown in Fig. 5(iv) which gives logic high input to SMFP-LD3 and logic zero input to SMFP-LD4. With this condition, 10 Gb/s input data which is modulated by TL3 is outputted to output port 2. In this case, the data signal inputted to SMFP-LD4 is not suppressed since the injection locking phenomena does not occurred with the signal λ2 as it is logic “0”. The reverse phenomenon occurs at the output port 1 with no data signal since there is the presence of input beam to SMFP-LD3, λ1, which suppresses the input data signal of 10 Gb/s. Both output ports will maintain the same state even with the removal of set beam because the power of λ1 is sufficient enough to maintain the suppression of dominant mode of SMFP-LD2 due to the presence of dominant mode of SMFP-LD1 as shown in Fig. 5(a)-5(v).
Fig. 5 Spectrum diagram for the control unit of proposed short pulse controlled optical switch: SR latch. (a) For the set state and data output at port 2 (b) for reset state and data output at port 1.
We found that with 0.12 nm wavelength detuning of λ1 to the corresponding side mode of SMFP-LD2, the minimum power needed to suppress the dominant mode is about −9 dBm (measured after coupler) and −12 dBm is required to maintain in the locked-state. Hence the self-injected mode of SMFP-LD1 should be less than −9 dBm but higher than −12 dBm. With control inputs S = 1 and R = 0, the data inputted to SMFP-LD3 will be suppressed and hence no data output is obtained at the output port 1 but the inputted data to SMFP-LD4 will be outputted to port 2 as shown in Fig. 6(a). This state will be maintained even later when S = 0 and R = 0, until R = 1 since the control signals, which are used to control the switching units, are the outputs from the S and R latch. Figure 6(a) includes spectrum diagram of the SMFP-LD4 with dominant mode at the wavelength of 1544.2 nm [Fig. 6(a)-6(i)], which is suppressed due to the presence of input data signal at the wavelength of 1551.02 nm [Fig. 6(a)-6(ii)]. Since there is an absence of λ2, data will be obtained at port 2. In presence of λ2 (when S = 0 and R = 1) obtained from BPF3, gain of the data signal can be modulated by λ2, which is the control signal for switching the data to output Port 2. The contrast ratio of 28.04 dB is measured for the output port 2 [Fig. 6(a)-6(iii)].
Fig. 6 Spectrum diagram for the switching unit of proposed short pulse controlled optical switch (a) port 2 (b) port 1.
Similarly, the output of Port 1 is obtained from SMFP-LD3 after passing through BPF2. The dominant mode of SMFP-LD3, 1539.84 nm, [Fig. 6(b)-6(i)] is suppressed by the injection of data signal at the wavelength of 1551.02 nm [Fig. 6(b)-6(ii)]. When reset control beam is logic 1, both inputs to SMFP-LD2, set beam and dominant mode of SMFP-LD1 are logic zero as the dominant mode of SMFP-LD1 is suppressed due to the presence of reset beam and S is set to “0”. Hence SMFP-LD2 gives high output at the wavelength of λ2, as shown in Fig. 5(b)-5(i), because both the inputs to SMFP-LD2 are logic 0. The output of SMFP-LD2 has the output of λ2 beam which suppresses input data at the output port 2 as shown in Fig. 6(b)-6(iii) whereas there will be the presence of data output at output port 1 since the beam injected to SMFP-LD3, λ1 is logic 0. The data will be outputted to the output port 1 even latter when R = 0 and S = 0 as shown in Fig. 5(b)-5(ii) and Fig. 6; until the set beam is “1” with reset beam is “0”.
Figure 7 shows oscilloscope traces at different stages of the proposed short pulse controlled optical switch. PPG1 and PPG2 are arranged to generate different duration of set and reset pulse, and accordingly the output state of Q and changes which are the control signals for controlling input data to be blocked or passed through output ports. We can see that even when the set or reset pulse is made logic 0 after the logic 1 state, the control signal for switch, Q and maintain the same state.
Fig. 7 Waveform diagram of input control signals, output of control units, input data, switched output data, rising/falling time of outputs and eye daigram of proposed short pulse controlled switch with 500 ps/div, 50 ps/div and 20 ps /div, respectively.
For example, even though the set pulse is “0” with previous state of S = 1 and R = 0, the logic state of Q and is maintained with logic state 1 and 0 respectively and vice-versa which controls the data transfer to output ports. When S beam is logic 1, 10 Gb/s NRZ input data is switched at the output port 2 and it maintains same state until the reset beam is logic 1. This shows the proposed optical switch can perform well even with a short pulse of control signals S and R. The minimum pulse duration used in the experiment is 280 ps for input S and R. Clear waveforms with rising/falling time of 31.6 ps/35.2 ps for output port 1 and 36 ps/41.7 ps for output port 2 with 10 Gbps 16 bit NRZ signal, eye diagrams with an extinction ratio of 12.33 dB for output port 1 and 11.47 dB for output port 2 with 10 Gb/s 231 − 1 pseudo random bit sequence input data signal are observed. Figure 8 shows no noise floor up to BER of 10−12 and maximum power penalty of about 1.7 dB is measured at the BER of 10−9. The clear output waveforms, rising/falling time, eye diagrams, and BER plot show the good performance of proposed short pulse controlled optical switch and proves the proposed switch using single mode Fabry-Pérot laser diodes.
4. Conclusion
We have proposed and demonstrated a novel approach of short pulse controlled optical switch using self-injected single mode Fabry-Pérot laser diode along with the analysis of wavelength detuning effect on hysteresis width and rising/falling time of SMFP-LD. The analysis provides the information for choosing the wavelength detuning and hence the optimal power for the different beams to be used in the proposed idea. The minimum pulse duration of 280 ps are used for switching 10 Gb/s input data at the output ports. The output of the proposed scheme is verified with clear waveforms, eye diagrams with extinction ratio of more than 11 dB for both output ports, less rising/falling time of about 30 ps and 40 ps, and maximum power penalty of about 1.7 dB at BER of 10−9. Since SMFP-LDs are used in the proposed scheme, there is no need of additional probe beam and high biasing current as required in SOAs and other methods. Besides SMFP-LDs cost less than other active device used for the switching purpose which we believe is cost effective and power efficient. The maximum input data rate of 10 Gb/s can be further increased by increasing the relaxation oscillation of solitary lasers, which can be obtained by engineering the cavity length, the injected power and the injected current. The demonstrated short pulse optical switch can be used for power saving architecture and controlling applications, where there is a need to control the output with pulses. The calculation of total power consumption and its footprint on power saving along with the integration issues of device is already under research, which will provide more benefits to the applications in optical networks with short pulse controlled optical switch using SMFP-LDs.
References and links
1. Y. Liu, E. Tangdiongga, Z. Li, H. de Waardt, A. M. J. Koonen, G. D. Khoe, X. Shu, I. Bennion, and H. J. S. Dorren, “Error-free 320-Gb/s all-optical wavelength conversion using a single semiconductor optical amplifier,” J. Opt. Soc. Am. B 25, 103–108 (2007).
2. E. Tangdiongga, H. C. H. Mulvad, H. de Waardt, G. D. Khoe, A. M. J. Koonen, and H. J. S. Dorren, “SOA- based clock recovery and demultiplexing in a lab trial of 640-Gb/s OTDM transmission over 50-km fibre link,” Proceedings of European Conference and Exhibition on Optical Communication, (Berlin, 2007), pp. 57–58.
3. J. T. Kim, J. J. Ju, S. Park, M. S. Kim, S. K. Park, and M. H. Lee, “Chip-to-chip optical interconnect using gold long-range surface plasmon polariton waveguides,” Opt. Express 16(17), 13133–13138 (2008). [CrossRef] [PubMed]
4. K. J. Barker, A. Benner, R. Hoare, A. Hoisie, A. K. Jones, D. K. Kerbyson, D. Li, R. Melhem, R. Rajamony, E. Schenfeld, S. Shao, C. Stunkel, and P. Walker, “On the feasibility of optical circuit switching for high performance computing systems,” Proceedings of ACM/IEEE Supercomputing, (Seattle, WA, 2005), pp. 16. [CrossRef]
5. R. Hemenway, R. Grzybowski, C. Minkenberg, and R. Luijten, “Optical packet-switched interconnect for super-computer applications,” J. Opt. Netw. 3(12), 900–914 (2004). [CrossRef]
6. D. M. Gvozdić, M. M. Krstić, and J. V. Crnjanski, “Switching time in optically bistable injection-locked semiconductor lasers,” Opt. Lett. 36(21), 4200–4202 (2011). [CrossRef] [PubMed]
7. M. M. Krstić, J. V. Crnjanski, and D. M. Gvozdić, “Injection power and detuning-dependent bistability in Fabry Pérot laser diodes,” IEEE J. Quantum Electron. 18(2), 826–833 (2012). [CrossRef]
8. G. T. Kanellos, D. Fitsios, T. Alexoudi, C. Vagionas, A. Miliou, and N. Pleros, “Bringing WDM Into Optical Static RAM Architectures,” J. Lightwave Technol. 31(6), 988–995 (2013). [CrossRef]
9. P. Maniotis, D. Fitsios, G. T. Kanellos, and N. Pleros, “Optical Buffering for Chip Multiprocessors: A 16GHz Optical Cache Memory Architecture,” J. Lightwave Technol. 31(24), 4175–4191 (2013). [CrossRef]
10. Z. Zang and Y. Zhang, “Low-switching power (<45mW) optical bistability based on optical nonlinearity of ytterbium-doped fiber with a fiber Bragg grating pair,” J. Mod. Opt. 59(2), 161–165 (2012). [CrossRef]
11. M. Jamali, V. Ahmadi, and M. Razaghi, “Optical self-switching based on a semiconductor- optical-amplifier-assisted Sagnac interferometer,” J. Opt. Soc. Am. B 30(10), 2576–2583 (2013). [CrossRef]
12. H. Ju, S. Zhang, D. Lenstra, H. de Waardt, E. Tangdiongga, G. Khoe, and H. J. S. Dorren, “SOA-based all-optical switch with subpicosecond full recovery,” Opt. Express 13(3), 942–947 (2005). [CrossRef] [PubMed]
13. Z. Zang and W. Yang, “Theoretical and experimental investigation of all-optical switching based on cascaded LPFGs separated by an erbium-doped fiber,” J. Appl. Opt. 109, 103106 (2011).
14. Z. Zang and Y. Zhang, “Analysis of optical switching in a Yb3+-doped fiber Bragg grating by using self-phase modulation and cross-phase modulation,” Appl. Opt. 51(16), 3424–3430 (2012). [CrossRef] [PubMed]
15. J. M. Martinez, J. Herrera, F. Ramos, and J. Marti, “All-optical Address Recognition Scheme for Label Swapping Networks,” IEEE Photon. Technol. Lett. 18(1), 151–153 (2006). [CrossRef]
16. J. P. Wang, B. S. Robinson, S. J. Savage, S. A. Hamilton, E. P. Ippen, R. Mu, H. Wang, J. Sarathy, and B. B. Stefanov, “Efficient performance optimization of SOA-MZI devices,” Opt. Express 16(5), 3288–3292 (2008). [CrossRef] [PubMed]
17. L. Y. Chan, P. K. A. Wai, L. F. K. Lui, B. Moses, W. H. Chung, H. Y. Tam, and M. S. Demokan, “Demonstration of an all-optical switch by use of a multiwavelength mutual injection-locked laser diode,” Opt. Lett. 28(10), 837–839 (2003). [CrossRef] [PubMed]
18. M. Uddin, J. S. Cho, and Y. H. Won, “All-optical multicasting NOT and NOR logic gates using gain modulation in an FP-LD,” IEICE Electron. Express 6(2), 104–110 (2009). [CrossRef]
19. L. Y. Chan, K. K. Qureshi, P. K. A. Wai, B. Moses, L. F. K. Lui, H. Y. Tam, and M. S. Demokan, “All-optical bit-error monitoring system using cascaded inverted wavelength converter and optical NOR gate,” IEEE Photon. Technol. Lett. 15(4), 593–595 (2003). [CrossRef]
20. Y. D. Jeong, J. S. Cho, Y. H. Won, H. J. Lee, and H. Yoo, “All-optical flip-flop based on the bistability of injection locked Fabry-Perot laser diode,” Opt. Express 14(9), 4058–4063 (2006). [CrossRef] [PubMed]
21. B. Nakarmi, M. Rakib-Uddin, T. Q. Hoai, and Y. H. Won, “A simple controlled all-optical ON/OFF switch using gain modulation in single mode Fabry-Prot laser diode,” IEEE Photon. Technol. Lett. 24, 212–214 (2011). [CrossRef]
22. B. Nakarmi, M. Rakib-Uddin, and Y. H. Won, “Realization of all-optical multi-logic functions and a digital adder with input beam power management for multi-input injection locking in a single-mode Fabry-Pérot laser diode,” Opt. Express 19(15), 14121–14129 (2011). [CrossRef] [PubMed]
23. B. Nakarmi, Y. H. Won, and X. Zhang, “External cavity based single mode Fabry-Prot laser diode and its application towards all-optical digital circuits,” Proceedings of SPIE Photonics Asia, (Beijing, 2012), pp. 57–58.
24. E. K. Lau, H. K. Sung, and M. C. Wu, “Frequency response enhancement of optical injection-locked lasers,” IEEE J. Quantum Electron. 44(1), 90–99 (2008). [CrossRef]
25. Y. D. Jeong, Y. H. Won, S. O. Choi, and J. H. Yoon, “Tunable single-mode Fabry-Perot laser diode using a built-in external cavity and its modulation characteristics,” Opt. Lett. 31(17), 2586–2588 (2006). [CrossRef] [PubMed]
26. M. Krstić, J. Crnjanski, M. Masšanović, L. Johansson, L. Coldren, and D. Gvozdić, “Multivalued stability map of an injection-locked semiconductor laser,” IEEE J. Quantum Electron. 19(4), 1501408 (2013). [CrossRef]
27. H. Yoo, Y. D. Jeong, Y. H. Won, M. Kang, and H. J. Lee, “All-optical wavelength conversion using absorption modulation of an injection-locked Fabry-Pérot laser diode,” IEEE Photon. Technol. Lett. 16(2), 536–538 (2004). [CrossRef]
