Preview only show first 10 pages with watermark. For full document please download

An Efficient Time Lens For Optical Pulse Compression

   EMBED


Share

Transcript

An Efficient time lens for optical pulse compression Taraprasad Chattopadhyay and Sujit Das Presented by: Prof. Taraprasad Chattopadhyay Department of Physics Visva-Bharati University Santiniketan, West Bengal, India. Why do we need optical pulse compression? • Conventional methods generate optical pulses having widths in the picosec. region. • Femto second pulses find applications in optical time division multiplexed (OTDM) communication system. OTDM is a high capacity, high bit-rate communication system. • High bit-rate requires high repetition rate of the pulse. • To generate high repetition rate pulses we need ultra-short widths of optical pulses. • This is one application – there are other applications as well. How to generate ultra-short optical pulses? • Normal pulses can be compressed in the time domain Generation of ultra-short optical pulses How can we produce pulse compression? by using a time lens What is a time lens? • Ordinary lenses used in geometrical optics are spatial lenses. They compress a light beam in space. • On the other hand, a time lens compresses the light pulse in time reduction of optical pulse width How to design a time lens ? • Apply phase modulation of the pulse and then pass through a single-mode fiber to exploit group velocity dispersion. • Phase modulation broadens the spectrum of the pulse, while GVD produces a cancellation of pulse broadening. overall pulse compression Author’s proposal • Use a cascade of phase modulators instead of a single optical phase modulator in the design of the time lens. • Consider the same total drive power for the sake of comparison of the cascade with a single modulator. Proposed time lens produces better pulse compression. Consider a cascade of N identical optical phase modulators. The phase modulation produced is N Δϕ(t)=(π⁄Vπ)∑Vn (t) n=1 …..(1) Vπ= half-wave voltage of the constituent phase modulators. Take Vn(t) = Vm0 cos ωmt ….(2) for all ‘N’. Vm0 = modulating signal voltage, and ωm = modulator drive frequency Case I. Linear chirping of the pulse Let τd = input pulse half-width Assume ωmτd <<1. Then, cosωmt≈ [1- ωm2t2 /2] ……(3) Assume Gaussian electric field profile of the input pulse. Then, Ein (t) =Ep exp[-t2 ln2/(2τd2 )] ….(4) The electric field of the light pulse at the output of the cascade modulator is given by Ein(t) =Ep exp[-t2 ln2/(2τd2 )].exp[-j(φ0 –NFt2 )] …(5) Where φ0 =(πVm0⁄Vπ)N F=(πVm0⁄Vπ) ωm2 /2 ….(6) …(7) Time-dependent frequency variation of the pulse is Δω(t)=dΔφ(t)/dt =2NFt ….(8) Linear chirping When the chirped pulse travels through a single mode fiber it undergoes spreading due to dispersion. It has a half width given by • • • • • • ΔτL =[d(L/vg ) /dω]Δω = Lβ2Δω ….(9) where Δω = 2NFτd β2 = d(1/vg ) /dω = d2 βp / dω2 vg = group velocity of the light pulse β2 = GVD parameter βp= phase shift constant of the lightwave in the fiber Dispersion parameter of the single mode fiber is related with GVD parameter as • β2 =-λ2D/2πc ……(11) • λ = vacuum wavelength of light • c= vacuum velocity of light In presence of linear chirping, the resultant half width of the pulse is τR=τd +ΔτL =τd(1+2Lβ2NF) ….(12) pulse compression since β2 <0 in standard single-mode fiber. Theoretically, τR 0 when 2Lβ2NF 1. The optimum fiber length is given by • Lopt=1/(2NF|β2|)=1/[N(πVm0⁄Vπ) ωm2 |β2|] ….(13) • Now, Pin=NPm and Vm0 ∞√Pm = √(Pin/N) ….(14) • Here, Pin= total drive power input to the cascade • Pm= drive power input to the single constituent modulator. • Eqn.(13) • Lopt is reduced by √N. One advantage of the cascade Eqn.(1) • Δφ(t) is increased √N times in the cascade. Another advantage of the cascade. • Taking λ=1.55 μm , D=17 ps/(Km-nm), fm=10 GHz, Vm0⁄Vπ=0.5, we get • Lopt=7.442 Km for N=1, and • Lopt=3.721 Km for N=4. • Taking τd=2.5 ps, fm=10 GHz, L=3.7Km, we get τR=14 fs. Nonlinear frequency chirping of the pulse • When ωmτd is not very small, we have • Δϕ(t)= N(πVm0⁄Vπ )[1- … ωm2t2/2 + ωm4t4/24] ...(15) • The resultant half-width of the pulse is • τR=τd[1+2Lβ2NF(1-ωm2τd2/12)] …(16) Conclusion: • Theory and design of an efficient time lens is proposed. • This time lens produces better optical pulse compression.