All-Optical Design of Switching Network using TOAD based Cross-bar Switch

DOI : 10.17577/IJERTCONV4IS28009

Download Full-Text PDF Cite this Publication

Text Only Version

All-Optical Design of Switching Network using TOAD based Cross-bar Switch

Ashis Kumar Mandal1, * , Goutam Kumar Maity2,3, Nabin Baran Manik1, Jonathon David White3

1Department of Physics, Jadavpur University, Kolkata-32, India

2Department of Electronics and Communication Engineering, MCKV Institute of Engineering, Howrah, India

3Department of Photonics Engineering, Yuan Ze University, Taoyuan, Taiwan

Abstract Recently operationally versatile semiconductor optical amplifier (SOA) based high speed optical switches in interferometric form are enormously important because of its own fast switching time, high repetition rate, low power consumption. This switch has contributed the revolution in all- optical information processing systems. Using this interferometric (2×2) all-optical Terahertz Optical Asymmetric Demultiplexer based switch; the switching network is designed to have four switching actions by the proposed circuit. Numerical simulation has been done to achieve the performance of the targeted design.

Keywords – Terahertz optical asymmetric demultiplexer; optical cross-bar switch; switching network.

  1. INTRODUCTION

    The opticalopticaloptical systems are established on optical switching frameworks. The dense wavelength division multiplexing of networks requires the all-optical switches [1]. These optical cross connect are significant system for their high bandwidth, lower power consumption, high connection density and low crosstalk [2]. For their speed and independence of data format, this optical switch has wide range of application in communication system [3, 4]. In the field of optical information processing systems the optical interferometric SOA-based switches are really good for fast switching time, and other factors. [5].

    The complex combinatorial circuits and systems can be formed using TOADs and the sequential circuits concerning the crucial issues of versatility, measurability and spatial property of circuits may be designed [6]. Due to so many advantages, the network of switch using all-optical cross-bar switch based on interferometric TOAD has been designed in this paper.

  2. ALL-OPTICAL (2×2) CROSS-BAR SWITCH

    The TOAD has capacity of demultiplexing data at 50 Gbits/s [7]. Also, the same group has reported that, by the decreasing the SOA length to 100 m and increasing of its dc bias current, its generation delay can be change up to 1 ps without impinging on its performance as a nonlinear element (NLE) then TOAD can perform Tb/s demultiplexing. The TOAD are made of a loop mirror with an additional intra- loop 2×2 coupler and this loop contains a control pulse (CP) of different wavelengths than 7that of incoming pulse and a se7miconductor optical amplifier (SOA) which is offset from the loops midpoint by a distance x [8].

    To design 2×2 optical cross-bar switch the circuit has been changed slightly. A and B are two inputs and P and Q are two outputs as displayed in Fig. 1. The input signals A and B are of different wavelength than CP that is introduced through a loop of 3-dB coupler. The output transmitted (Tr) and reflected (Re) transfer functions are [6, 9]:

    [CW means clockwise and CCW means counter clockwise]

    Tr (t) =1/4{Gcw(t) + Gccw(t) – 2 Gcw(t) Gccw(t).cos()} (1)

    Re (t) =1/4{Gcw(t) + Gccw(t) + 2 Gcw(t) Gccw(t).cos()} (2) The phase difference between cw and ccw pulse is defined by

    (cw ccw ) . The symbols Gcw (t),Gccw (t) indicate the

    2 G

    2 G

    respective power gains and ln Gcw , where

    ccw

    is line-width factor of enhancement.

    Fig. 1. TOAD based 2×2 cross-bar switch [FFilter and CROptical

    circulator]

    Fig. 2. Schematic diagram of TOAD based 2×2 cross-bar switch

    TABLE I. TRUTH TABLE OF CROSS-BAR SWITCH

    P= BC +AC

    A

    B

    CP

    P

    Q

    Remarks

    0

    0

    0

    0

    B

    0

    A

    Cross-state (×)

    0

    1

    0

    1

    0

    1

    0

    0

    0

    1

    1

    1

    0

    1

    1

    0

    0

    1

    0

    A

    0

    B

    Bar-state ()

    0

    1

    1

    0

    1

    1

    0

    1

    1

    0

    1

    1

    1

    1

    1

    A

    B

    CP

    P

    Q

    Remarks

    0

    0

    0

    0

    B

    0

    A

    Cross-state (×)

    0

    1

    0

    1

    0

    1

    0

    0

    0

    1

    1

    1

    0

    1

    1

    0

    0

    1

    0

    A

    0

    B

    Bar-state ()

    0

    1

    1

    0

    1

    1

    0

    1

    1

    0

    1

    1

    1

    1

    1

    Q= AC +BC (3)

    Hence, it is very easy to write, if C = 0 then P = B and Q = A cross-state (Fig. 3.) but, if C = 1 then P = A and Q = B bar-state (Fig. 4.).

  3. PROPOSED SWITCHING NETWORK

    The all-optical submitted switching network is formed by four cross-bar switches as shown in Fig. 5. There are four separate switching actions which are needed two control inputs. Here, the two control inputs are C1 and C2. For the four switching actions it is presumed the following:

    C1 C2 = 00 for the 1st switching actions (Fig. 6)

    nd

    C1 C2 = 01 for the 2 switching actions (Fig. 7)

    As a consequence, the two counter-propagation data signal will get a differential gain, i.e. Gccw Gcw. Hence they recombine at the input coupler, and the data pulse will come out from the transmitted port and Re (t) 0.

    Fig. 3. Schematic diagram of TOAD based 2×2 cross-bar switch when

    CP=0, i.e. Cross-state operation

    Fig. 4. 2×2 cross-bar switch when CP=1, i.e. Bar-state operation

    When the control signal is absent (i.e., CP = off), the incoming signal passes through the SOA at different times around the loop, and experience the same unsaturated amplifier gain G0 of SOA. The diagram of 2×2 cross-bar switch is shown in Fig. 2. The truth table of the switch is in Table I. From this table it is very easy to write [if CP C]

    C1 C2 = 10 for the 3rd switching actions (Fig. 8) C1 C2 = 11 for the 4th switching actions (Fig. 9)

    Now, X1, X2, X3 and X4 are inputs and Y1, Y2, Y3 and Y4 are outputs of the switching network. The following four equations for the above four switching actions are shown below.

    Y1 = C 2C 1X4 + C 2C1X2 + C2C 1X3 + C2C1X1 Y2 = C 2C 1X3 + C 2C1X1 + C2C 1X4 + C2C1X2 Y3 = C 2C 1X2 + C 2C1X4 + C2C 1X1 + C2C1X3

    Y4 = C 2C 1X1 + C 2C1X3 + C2C 1X2 + C2C1X4 (4)

    The above equations can be realized using the proposed design in fig. 5.

    Fig. 5. Design of proposed switching network, Inputs: X1, X2, X3 and X4, Outputs: Y1, Y2, Y3 and Y4 and Control inputs: C1 and C2

    Fig. 6. 1st Switching action

    Fig. 7. 2nd Switching action

    Fig. 8. 3rd Switching action

    Fig. 9. 4th Switching action

  4. LOGIC UNIT OF CROSS-BAR NETWORK

    The reconfigurable design by cross-bar switches is presented in Fig. 10. The switch-1 is connected to a constant signal. Signal A is the control pulse of switch 1. The outputs of it are A and A . They are inputs of switch 2 and switch 3 respectively [Fig. 10]. And B is the control pulse of switch 2 and switch 3. Hence AB , AB, A B and A B are fed to the input of switch-4, 5 and 6 of 3 x 3 cross-bar systems as shown in Fig. 10. The controls pulses switch 5 and switch 9 are 0. The control pulses of switch 4, 7, 8 and 6 are C4, C3, C2 and C1 respectively. We can change these control pulses to get different outputs as shown in Table II.

    TABLE II. THE BOOLEAN LOGICAL OPERATION OF FIG. 10

    Function

    C1

    C2

    C3

    C4

    Output

    f1

    0

    0

    0

    0

    A B

    f2

    0

    0

    0

    1

    A B

    f3

    0

    0

    1

    0

    AB

    f4

    0

    0

    1

    1

    0

    f5

    0

    1

    0

    0

    A + B

    f6

    0

    1

    0

    1

    B

    f7

    0

    1

    1

    0

    A

    f8

    0

    1

    1

    1

    AB

    f9

    1

    0

    0

    0

    A + B

    f10

    1

    0

    0

    1

    A

    f11

    1

    0

    1

    0

    B

    f12

    1

    0

    1

    1

    A B

    Fig. 10. Reconfigurable design. BC Beam Combiner and CPLS Constant Pulsed Light Source

  5. SIMULATED RESULT

    Computer simulation (in Optiwave.OptiSystem.v7.0) has been done using the parameter [10, 11] of SOA, which is shown in Table III. Fig. 12 shows that X1, X2, X3 and X4 are inputs and Y1, Y2, Y3 and Y4 are outputs of Fig. 10. Here logic states of inputs (X1, X2, X3 and X4) and outputs (Y1, Y2, Y3 and Y4) are 1111, 1111, 0000, 1111, 1101, 1110,

    0111, 1011 respectively.

    TABLE III. PARAMETERS USED IN SIMULATION

    Parameters

    Symbol

    Value

    Injection current of SOA

    I

    250 mA

    Confinement factor

    0.5

    Differential gain

    aN

    3.3 x 10-20 m2

    Line-width enhancement factor of SOA

    7

    Carrier density at transparency

    Ntr

    1.0 x 1024 m-3

    Width of the active region of SOA

    w

    1.4 m

    Depth of the active region of SOA

    d

    300 nm

    Active length of SOA

    L

    140 m

    Internal loss of the wave guide

    D

    2500 m-1

    Wave length of light

    0

    1550 nm

    Gain recovery time

    e

    100 ps

    Control pulse energy

    Ec

    400fJ

    Full width at half maximum of control pulse

    2 ps

    Incoming pulse energy

    Ein

    0.01 mW

    Here, the section shows the simulation results that verify the all-optical proposed switching network. Fig. 12 is the simulation result of the proposed switching network.

    Inputs: X1, X2, and X4

    Input: X3

    Output: Y1

    Output: Y2

    Output: Y3

    Output: Y4

    Fig. 12. Inputs of switching network: X1, X2, X3 and X4 and outputs of switching network: Y1, Y2, Y3 and Y4

    In Fig. 11, when CP is ON with few sets of G0 the output transfer function (Tr of equation 1) versus time (ps) is drawn.

    And the expression of contrast ratio (C.R.) is given below [12].

    P1

    0

    0

    C.R.(dB) 10 log Min

    P

    (5)

    Max

    Fig. 11. Output transfer function versus time (ps)

    By eq. (5) we get the output C.R. (dB) for reversible logic gates is found 18.03 dB.

    Fig-13 shows give the variation of C.R. with control pulse energy with eccentricity of the loop is kept constant. The

    variation of C.R. with

    Ecp

    is shown in fig-13 and it

    confirms that C.R. is high when Ecp of the loop is 95.5 fJ.

    Fig. 13. Contrast ratio versus control pulse energy (Ec)

  6. CONCLUSION

In this paper, we have design of all-optical switching network using TOAD based cross-bar switches. The theoretical model is introduced in this paper and the simulation results also show that the proposed scheme achieves higher all-optical logic computing system. To get experimental result of this targeted design, some factors like, intensity losses due to beam splitters/fiber couplers, walk-of problem due to dispersion, optical circulator, etc are to be examined and the polarization controller may apply. We apply small length SOA and control pulse of very lower than 1 pJ to avoid the effect of amplified spontaneous emission. The future work would concentrate the realization of Digital Signal Processing.

REFERENCES

  1. S. Okamoto, A. Watanable, K-I. Sato, Optical path cross- connect node architectures for photonic transparent network, J. Lightw. Technol. 14 (6) (1996) 14101422.

  2. T. Chikama, H. Onaka, S. Kuroyanagi, Photonic networking using optical add drop multiplexer and optical cross-connects, FUJITSU Sci. Tech. J. 35 (1) (1999) 4655.

  3. H. Obara, Extended class of reduced crossbar switches, in: The 2009 international Conf. on advanced Tech. for Communication, IEEE, (2009) 129132, ISBN: 978-1-4244- 5139-5/09.

  4. L. Geo, J. Sun, X. Sun, T. Li, W. Gao, .Y. Yan, D. Zhang, Simulation and optimization of a polymer 2×2 multimode interference Mach-Zehnder interferometer electro optic switch with push pull electrodes, Opt. Laser Tech. 42 (2010) 8592.

  5. K.E. Zoiros, J. Vardakas, T. Houbavlis, M. Moyssidis, Investigation of SOAassisted Sagnac re-circulating shift register switching characteristics, Optik 116 (2005) 527541.

  6. K.E. Zoiros, R. Chasioti, C.S. Koukourlis, T. Houbavlis, On the output characteristics of a semiconductor optical amplifier driven by an ultrafast optical time division multiplexing pulse train, Optik 118 (3) (2007) 134146.

  7. J.P. Sokoloff, I. Glesk, P.R. Prucnal, R.K. Boneck, Performance of a 50 Gbit/s optical time domain multiplexed system using a terahertz optical asymmetric demultiplexer, IEEE Photon. Technol. Lett. 6 (1) (1994) 98100.

  8. A.J. Poustie, K.J. Blow, Demonstration of an all-optical Fredkin gate, Opt. Commun. 174 (2000) 317320.

  9. T. Chattopadhyay, All-optical symmetric ternary logic gate, Opt. Laser Technol. 42 (2010) 10141021.

  10. T. Chattopadhyay, All-optical cross-bar network architecture using TOAD based Interferometric switch and designing of reconfigurable logic unit, Optical Fiber Technology 17 (2011) 558567.

  11. A.K. Mandal, G.K. Maity, All Optical Reversible NOR Gates using TOAD, International Journal of Computer Applications, 973-93-80883-42-5 (ISSN: 0975 8887), (2014) 18-23.

  12. Y. J. Jung, S. Lee, N. Park, All-optical 4-bit gray code to binary coded decimal converter, Optical Components and Materials, Proceedings of the SPIE, Volume 6890, 68900S, 2008.

Leave a Reply