CHARACTERISATION OF A NOVEL DUAL-CONTROL TOAD SWITCH

1. CHARACTERISATION OF A NOVEL DUAL-CONTROL TOAD SWITCH H Le-Minh, Z Ghassemlooy, and W P Ng Optical Communications Research Group School of Informatics, Engineering & Technology Northumbria University, Newcastle, UK Lancaster, 30/03 – 01/04/2005

2. Outlines • Introduction • All-optical switches • TOAD switch: single & dual control • Numerical modeling of SOA • Simulation Results • Conclusions

3. Introduction • To enhance high-capacity optical network • Multiplexing: DWDM and OTDM • Higher channel capacity (higher aggregate bit rate) • All optical switching: • Optical transparency: removing O-E-O conversions Need an ultra-fast all-optical switches

4. All-Optical Switches • Are based on: Nonlinear effect + optical interferometer • Configurations: • Nonlinear Optical Loop Mirror (NOLM) • Terahertz Optical Asymmetric Demultiplexer (TOAD) • Symmetric Mach-Zehnder (SMZ) • Ultrafast Nonlinear Interferometer (UNI)

5. SOA CW CCW Input Data Reflected Data a0o 0.707a0o 0.707a90o TOAD Switch • Short fibre loop as the optical interferometer: by the CW & CCW data components

6. Tasym SOA CW CCW Input Data saturates SOA Control pulse (CP) Transmitted Data TOAD switch • Short fibre loop as the optical interferometer: by the CW & CCW data components • Semiconductor Optical Amplifier (SOA) induces nonlinearity Switching window width is defined by the Tasym

7. SOA CW CCW Input Data Control pulse (CP) Transmitted Data Input Switched Output (Transmitted) TOAD switch • Short fibre loop (1 m) used as the optical interferometer: by the CW & CCW data components • Semiconductor Optical Amplifier (SOA): induces nonlinearity • Advantages • Possible to integrate in chip • Low control pulse (CP) energy • Disadvantages • Asymmetric switching window • High inter-channel crosstalk • Distorted signal pulse shape

8. 3 2 1 4 3 2 1 4 CP TOAD: Asymmetric Switching Window Single CP CW direction x 0 LSOA No effected by CP ( fully amplified after exiting SOA Follows CP ( experience full saturation effect after exiting SOA Same as pulse (3) if TSOA_recovery >> TSOA CCW direction CW direction 2 4 3 1 This pulse meets CP at x/2 ( experienced saturation effects of SOA segments up to x/2 LSOA – x Experienced more partial saturation effect than pulse (1) CCW direction 2 4 1 3 Experienced more partial saturation effect than pulses (1), (2) Any pulse following pulse (4) will experience the full saturation effect until SOA carrier density recovers

9. Single CP x 0 LSOA 2 1 3 3 4 1 4 2 CP CW direction Tasym 2 4 3 1 SW LSOA – x Gain GCW(t) CCW direction 2 4 1 3 GCCW(t) Time TOAD: Asymmetric Switching Window –contd. CW direction No effected by CP ( fully amplified after exiting SOA Follows CP ( experience full saturation effect after exiting SOA Same as pulse (3) if TSOA_recovery >> TSOA CCW direction This pulse meets CP at x/2 ( experienced saturation effects of SOA segments up to x/2 Experienced more partial saturation effect than pulse (1) Experienced more partial saturation effect than pulses (1), (2) Any pulse following pulse (4) will experience the full saturation effect until SOA carrier density recovers Reason: Difference of CW and CCW gain profiles and not steep

10. SW2 SW1 CW CCW Input Data CPCCW SOA Transmitted Data CPCW TOAD: Symmetric Switching Window • Cascading two TOAD switches (Prucnal’02) • Using dual-control in single TOAD switch • CPCW and CPCCW are identical • CPCW and CPCCW are simultaneously applied to the SOA • Therefore, CW and CCW data components will experience the same amplification & saturation effects • (GCW(t) and GCCW(t) are the same but delayed

11. Dual-CP x 0 LSOA LSOA – x 5 3 4 1 2 CPCCW CPCW - LSOA/2 LSOA + x 3 5 4 2 1 CW direction 1.5LSOA -x 2 3 5 1 4 CCW direction TOAD: Symmetric Switching Window with Dual Control Pulses CW direction Pulses before (1) do not meet CPCCW( experience full amplification Partial saturation by CPCCW More partial saturation by CPCCW If x<LSOA/2, affected by CPCW ( saturated by segments up to LSOA/2 If x>LSOA/2, segments from LSOA/2 to LSOA are further saturated by CPCW and CPCCW Pulses after (5) experience full double saturation of SOA when all CPs exit CCW direction The effects on CCW data pulses are exactlysame as in CW direction!

12. Dual-CP x 0 LSOA LSOA – x 5 4 2 1 3 CPCCW CPCW - LSOA/2 LSOA + x 3 5 4 2 1 CW direction Tasym SW 1.5LSOA -x 3 5 2 1 4 GCW(t) CCW direction GCCW(t) Time TOAD: Symmetric Switching Window with Dual Control Pulses CW direction Pulses before (1) do not meet CPCCW( experience full amplification Partial saturation by CPCCW More partial saturation by CPCCW If x<LSOA/2, affected by CPCW ( saturated by segments up to LSOA/2 If x>LSOA/2, segments from LSOA/2 to LSOA are further saturated by CPCW and CPCCW Pulses after (5) experience full double saturation of SOA when all CPs exit Gain CCW direction The effects on CCW data pulses are exactly the same as in CW direction!

13. k - 1 k + 1 k Modeling of SOA 1. SOA is divided into a number of small segments 2. At each segment, e.g. kth, the arriving powers are from CW & CCW directions 3. The carrier density at each segment is consequently updated by

14. Simulation Results I Gain profiles and switching windows Dual control:create the steep transitions in the temporal gain profiles ( help to create the steep switching window edges

15. Simulation Results II Carrier density in SOA when single control pulse going through Time angle

16. Simulation Results III SOA carrier density with both control pulses propagating within the SOA Single control Time angle

17. Simulation Results IV Dual control: induce less inter-channel crosstalk and less pulse-shape distortion of switched pulse

18. Conclusions • Using dual-control pulses in a TOAD configuration symmetric switching window profile is obtained • Inter-channel crosstalk and distortion of switched pulse are reduced

19. Thank you. Question please?