Fiber-Optic Communication via the Nonlinear Fourier Transform
|
|
- Ellen Sparks
- 5 years ago
- Views:
Transcription
1 Fiber-Optic Communication via the Nonlinear Fourier Transform Frank R. Kschischang Department of Electrical & Computer Engineering University of Toronto IC Research Day SwissTech Convention Center, EPFL 3 June 6
2 Thank You To: Mansoor Yousefi (University of Toronto, Technical University of Munich, Télécom ParisTech) Gerhard Kramer (Technical University of Munich) Siddarth Hari (University of Toronto) Alan P.-T. Lau (Hong Kong Polytechnical University) René-Jean Essiambre (Nokia Bell Labs) The organizers: Michael Gastpar Rüdiger Urbanke
3 Part I: Fiber-Optic Communication 3
4 Illustrating the principle of total internal reflection.
5 Source: Nicolas Rapp, Fortune Magazine, July 3, 5
6 Source: 6
7 eunetworks Optical Fiber Network Map Source: 7
8 Fiber-Optic Communication Systems Hybrid Amplifier SSMF Hybrid Amplifier ROADM SSMF Hybrid Amplifier ROADM SSMF Hybrid Amplifier TX RX Forward pump Backward pump Forward pump Backward pump Forward pump Backward pump span span span N Physics: Enabling Technologies Low-loss (. db/km) optical fiber, huge bandwidth ( 5 THz bandwidth) Optical amplifiers RX TX ROADM λ λ 3 Laser transmitters and Mach-Zehnder modulators TX RX 8
9 Historical Capacity Evolution System capacity (Tb/s) Pbit/s Tbit/s Tbit/s Tbit/s Gbit/s Gbit/s Gbit/s TDM Research WDM Research SDM Research Year Space-division multiplexing has already exceeded the nonlinear Shannon limit Nokia 6 < COPYRIGHT 6 Nokia. ALL RIGHTS RESERVED. > Nonlinear Shannon capacity limit estimate (8 Tb/s) Courtesy of R.-J. Essiambre (Nokia Bell Labs) [IEEE Photonics J., Apr. 3] TDM: Time-division multiplexing WDM: Wavelength-division multiplexing SDM: Space-division multiplexing 9
10 Nonlinear Shannon Limit Estimate (Single Polarization) 5 km of standard single-mode fiber (SSMF) Spectral eficiency (bits/s/hz) Shannon AWGN Record laboratory experiment s Nonlinear Shannon limit of spectral efficiency Courtesy of R.-J. Essiambre (Nokia Bell Labs) [J. Lightwave Techn., Feb. ] Signal-to-noise ratio, SNR* (db) *SNR is the signal average power average divided by the AWGN power per symbol Experimentally demonstrated fiber capacity have leveled off to ~6% of maximum spectral efficiency since Nokia 6 < COPYRIGHT 6 Nokia. ALL RIGHTS RESERVED. >
11 Fiber-Optic Communication Systems: Challenges Reliability P e < 5
12 Fiber-Optic Communication Systems: Challenges Reliability P e < 5 Speed Gb/s per-channel data rates (or even more)
13 Fiber-Optic Communication Systems: Challenges Reliability P e < 5 Speed Gb/s per-channel data rates (or even more) Non-Linearity The fiber-optic channel is non-linear in the input power!
14 The Kerr Effect Kerr electro-optic effect (DC Kerr effect) An effect discovered by John Kerr in 875 It produces a change of refractive index in the direction parallel to the externally applied electric field The change of index is proportional to the square of the magnitude of the external field
15 The Kerr Effect in Optical Fibers Optical Kerr Effect (or AC Kerr effect) No externally applied electric field is necessary The signal light itself produces the electric field that changes the index of refraction of the material (fused silica) The change in index in turn changes the signal field The change in index of refraction is proportional to the square of the field magnitude changes index of refraction optical fiber P WDM channel generates distortion channel of interest f 3
16 Nonlinear Effects in Fibers intra-channel inter-channel signal noise signal signal signal noise signal signal NL phase noise parametric amplification self-phase modulation NL phase noise WDM nonlinearities SPM-induced NL phase noise modulation instability isolated pulse SPM IXPM IFWM XPM-induced NL phase noise XPM FWM Key NL = nonlinear; SPM = self-phase modulation; (I)XPM = (intra-channel) cross-phase modulation; (I)FWM = (intra-channel) four-wave mixing; WDM = wavelength-division multiplexing Courtesy of R. J. Essiambre
17 System Model Hybrid Amplifier coherent fiber-optic communication system standard-single-mode fiber ideal distributed Raman amplification SSMF Hybrid Amplifier ROADM SSMF Hybrid Amplifier ROADM SSMF Hybrid Amplifier TX RX Forward pump Backward pump Forward pump Backward pump Forward pump Backward pump span span span N But, could also consider systems with inline dispersion-compensating fiber lumped amplification Q(t, ) Q(t, L) TX AMP AMP RX SSMF }{{} N spans 5
18 Generalized Nonlinear Schrödinger Equation Q(t, z) is the complex baseband representation of the signal (the full field, representing co-propagating DWDM signals) Transmitter sends Q(t, ) Receiver gets Q(t, L), where L is the total system length Evolution of Q(t, z) The generalized non-linear Schrödinger (GNLS) equation expresses the evolution of Q(t, z): Q(t, z) z + jβ Q(t, z) t }{{} dispersion jγ Q(t, z) Q(t, z) } {{ } nonlinearity = n(t, z) }{{} noise No loss term (since ideal distributed Raman amplification is assumed). 6
19 System Parameters n(t, z) is a circularly symmetric complex Gaussian noise process with autocorrelation E [ n(t, z)n (t, z ) ] = αhν s K T δ(t t, z z ), where h is Planck s constant, ν s is the optical frequency, and K T is the phonon occupancy factor. Second-order dispersion β ps /km Loss α.65 5 m Nonlinear coefficient γ.7 W km Center carrier frequency ν s 93. THz Phonon occupancy factor K T.3 7
20 Solving the GNLS Equation Throughout propagation over an optical fiber, stochastic effects (noise), linear effects (dispersion) and nonlinear effects (Kerr nonlinearity) interact. noise nonlinearity dispersion Even in the absence of noise, solving the GNLS equation requires numerical techniques. 8
21 Split-Step Fourier Method h divide fiber length into short segments consider each segment as the concatenation of (separable) nonlinear and linear transforms for distributed amplification, an additive noise is added after the linear step. Q(t, z ) Q(t, z + h) step size h 9
22 Nonlinear Step In the absence of linear effects, the GNLS equation has the form Q(t, z) z = jγ Q(t, z) Q(t, z), with solution Q(t, z + h) = Q(t, z ) exp(jγ Q(t, z ) h). Nonlinear Step is best taken in the time domain.
23 Linear Step Now use the previous solution as input to the linear step, i.e., let ˆQ(t, z ) = Q(t, z ) exp(jγ Q(t, z ) h) be the input to the linear step. The linear form of the GNLS equation is Q(t, z) = jβ Q(t, z) z t. Defining Q(t, z) = π it can be shown that Linear Step... Q(ω, z) exp(jωt)dω, i.e., Q(t, z) Q(ω, z + h) = Q(ω, z ) exp... is best taken in the frequency domain. ( j β ) ω h. F Q(ω, z),
24 Split-Step Propagator Putting this together, we have Q(t, z + h) = F F Q(t, z ) exp(jγ Q(t, z ) h) }{{} exp nonlinear step ( j β ) ω h } {{ } linear step where F is the Fourier transform operator. In practice: extensive use is made of the FFT step size can be adapted 3 linear and nonlinear steps can be offset by a half-step See: Sinkin, Holzlohner, Zweck, Menyuk, J. Lightwave Tech., 3.
25 Launching a Pulse z = km 8 q(t) t (soliton units) 3
26 Launching a Pulse z = km 8 q(t) t (soliton units) 3
27 Launching a Pulse z = 5 km 8 q(t) t (soliton units) 3
28 Launching a Pulse z = km 8 q(t) t (soliton units) 3
29 Current Approaches Digital Backpropagation q(t, ) q(t, L ) = K NLS (q(t, )) q(t, L ) = K NLS (q(t, L q(t, L) )) Wavelength-division multiplexing (WDM) input output noise W guard band }{{} out of band }{{} COI f
30 Digital Backpropagation Digital backpropagation = split-step Fourier method, using a negative step-size h, performed at the receiver Full compensation (involving multiple WDM channels) generally impossible, even in absence of noise (due to wavelength routing) Noise is neglected (cf. zero-forcing equalizer) Single-channel backpropagation typically performed, after extraction of desired channel using a filter 5
31 Current Achievable Rates pulse shapes constellation I(x) R(x) Amplitude time 6
32 Capacity Estimation See: R.-J. Essiambre, G. Kramer, P. J. Winzer, G. J. Foschini, B. Goebel, Capacity limits of optical fiber networks, J. Lightw. Technol., vol. 8, pp. 66 7, Sept./Oct.. System Model ( ) T s Q(t, ) Q(t, L) t = kt s TX CHANNEL LPF DSP 7
33 Achievable Rates from Memoryless Capacity Estimate 9 Spectral Efficiency (bits/s/hz) L= km, Eq. only L= km, BP L= km, Eq. only L= km, BP L=5 km, Eq. only L=5 km, BP Shannon Limit (AWGN) (BP adds.55 to.75 bits/s/hz relative to EQ) SNR (db) B. P. Smith and F. R. Kschischang, J. Lightwave Techn., vol. 3, pp. 7 53,. 8
34 Schematic Achievable Rate Curve 5 Rate [bits/s/hz] 5 log( + SNR) Nonlinearity kicks in SNR [db] noise-limited nonlinearity-limited 9
35 Optical Fiber Capacity Bounds C [bit/symbol] upper bound lower bounds power bound Upper bound: Transmit power G. Kramer, M. I. Yousefi, and F. R. Kschischang, Upper Bound on the Capacity of a Cascade of Nonlinear and Noisy Channels, Proc. IEEE Info. Theory Workshop, Jerusalem, Israel, Apr. 5. M. I. Yousefi, G. Kramer, and F. R. Kschischang, Upper Bound on the Capacity of the Nonlinear Schrödinger Channel, Proc. th Canadian Workshop on Info. Theory, St. John s, NL, Jul. 5. 3
36 Summary Pulse propagation in optical fibers is well-modeled by the stochastic nonlinear Schrödinger equation: Q(t, z) z + jβ Q(t, z) t }{{} dispersion jγ Q(t, z) Q(t, z) } {{ } nonlinearity = n(t, z) }{{} noise z is distance along the fiber; TX at z =, RX at z = L t is retarded time, i.e., t = τ β z where τ is ordinary time and β is a constant Q(t, z) is the complex envelope of the propagating signal β is the chromatic dispersion coefficient γ is the nonlinearity parameter V (t, z) is space-time white Gaussian noise noise dispersion nonlinearity 3
37 Part II: Nonlinear Fourier Transform b a b a NFT(q)(λ) 5 spectrum of L Rλ 5 Iλ 3
38 Central Question: Does fiber nonlinearity really place an upper limit on achievable spectral efficiency? To try to answer this question, we have to understand the nonlinearity more deeply. Spoiler: We will not give a satisfactory answer in this talk. 33
39 Evolution Equations An evolution equation (in + dimensions) is a partial differential equation for an unknown function q(t, z) of the form q z = K(q), where K(q) is an expression involving only q and its derivatives with respect to t. Heat: K(q) = c q tt NLSE: K(q) = j ( q tt (t, z) + q(t, z) q(t, z) ) KdV: K(q) = q ttt (t, z) + q(t, z)q t (t, z) (Here subscripts denote partial derivatives, thus q tt = t q(t, z).) 3
40 Normalized NLS Equation Changing variables (to so-called soliton units ): q = Q P, z = z L, t = t T, with T = we get the: β L and P = Normalized NLS equation γl (and then dropping the primes ) jq z (t, z) = q tt + q(t, z) q(t, z) + v(t, z). 35
41 Isospectral Flow A Key Idea We seek an invariant under evolution (in the absence of noise), e.g., let L be a linear differential operator (depending on q(t, z)). It may be possible to find an L whose (eigenvalue) spectrum remains constant, even as q evolves (in z) according to some evolution equation. z L q(t, ) q(t, z) q(t, L) L(q(t, )) L(q(t, z)) L(q(t, L)) Constant Spectrum 36
42 Example: Isospectral Families of Matrices Let L(z) be a family of diagonalizable square matrices whose entries are functions of z. Clearly, the eigenvalues of these matrices in general depend on z. For some families (isospectral families), it might be the case that while the entries of the matrix change with z, the eigenvalues remain constant. Each member of an isospectral family is similar to a constant diagonal matrix Λ, i.e., L(z) = G(z)ΛG (z) for some similarity transformation G(z). 37
43 Isospectral Families of Operators Let H be a Hilbert space, and let L(z) be a family of diagonalizable bounded linear operators operating on H, e.g., L(q(t, z)) = + q(t, z). t If the eigenvalues of L(z) do not depend on z, then we refer to L(z) as an isospectral family of operators. For each z, L(z) is similar to a multiplication operator Λ (the operator equivalent of a diagonal matrix), i.e., L(z) = G(z)ΛG (z), for some operator G(z). The spectrum of an operator is defined as σ(l) = {λ L λi is not invertible}. Spectrum can be discrete (like matrices), continuous, residual, etc. 38
44 The Lax Equation We have L(z) = G(z)ΛG (z), where Λ does not depend on z. Assuming that L(z) varies smoothly with z, we can form dl(z) dz = G ΛG + GΛ ( G G G ) = G } {{ G ( } GΛG ) ( GΛG ) G }{{}}{{}} {{ G } M(z) L(z) L(z) M(z) = M(z)L(z) L(z)M(z) = [M, L] () where [M, L] = ML LM is the commutator bracket. In other words, every diagonalizable isospectral operator L(z) satisfies the differential equation (). The converse is also true. 39
45 Lax Pairs Lemma Let L(z) be a diagonalizable family of operators. Then L(z) is an isospectral family if and only if it satisfies dl dz for some operator M, where [M, L] = ML LM. = [M, L], () Definition The operators L and M satisfying () are called a Lax Pair (after Peter D. Lax, who introduced the concept [968]).
46 Lax Pairs Recall that L may depend on a function q(t, z) and so can M. The commutator bracket [M, L] can create nonlinear evolution equations for q(t, z) in the form L z = [L, M] q z = K(q), where K(q) is some, in general nonlinear, function of q(t, z) and its time derivatives.
47 KdV Equation An example of this is the Korteweg-de Vries (KdV) equation (which arises in the evolution of waves on shallow water surfaces). Let q(t, z) be a real-valued function and choose L = t + 3 q, M = 3 t 3 + q t + q t. The Lax equation L z = [M, L] is easily simplified to 3 q z 3 (q ttt + qq t ) + (some terms) t, where is the zero operator. The zero-order term of this equation, which must be zero, produces the KdV equation q z = q ttt + qq t.
48 NLS equation If q(t, z) varies based on the NLS equation jq z = q tt + q q then the spectrum of L = j ) t q(t, z) q (t, z) t ( (3) is preserved during the evolution (Zakharov-Shabat 97). In addition, the operator M is given by ( jλ M = j q(t, z) ) λq(t, z) jq t (t, z) λq (t, z) jqt (t, z) jλ + j q(t, z) Thus the NLS equation is indeed generated by a Lax pair! 3
49 Eigenvalues and Eigenvectors of L The eigenvalues of the operator L, which are constant in an isospectral flow, are defined via Lv = λv () Taking the z derivative of () and using the Lax equation L z = [M, L], we can show that an eigenvector of L evolves according to the linear equation v z = Mv. Furthermore, we can re-write () as v t = Pv for some operator P. Thus we get Evolution of Eigenvectors of L The eigenvectors of L evolve according to the linear system v z = Mv (5) v t = Pv (6)
50 Zero-Curvature Condition Combining equations (5) and (6) by using the equality of mixed derivatives, i.e., v tz = v zt, the Lax equation () is reduced to the Zero-curvature Condition P z M t + [P, M] =. (7) Note that the nonlinear equation derived from (7) results as a compatibility condition between the two linear equations (5) and (6). Thus certain nonlinear evolution equations possess a certain hidden linearity. We can work with the Lax pair (L, M) or equivalently with the pair (P, M). 5
51 Integrable Systems We refer to a system described by a Lax pair M, L, measured at distance L, as the integrable system S = (L, M; L). x(t) = q(t, ) y(t) = q(t, L) L z (q) = [M(q), L(q)] input waveform output waveform Definition (Lax Convolution) The action of an integrable system S = (L, M; L) on the input q(t, ) is called the Lax convolution of q with S. We write the system output as q(t, L) = q(t, ) (L, M; L). Definition (Integrable communication channels) A waveform communication channel C : x(t) v(t, z) y(t) with inputs x(t) L (R) and space-time noise v(t, z) L (R, R + ), and output y(t) L (R), is said to be integrable if the noise-free channel is an integrable system. 6
52 Spectrum of L The Zakharov-Shabat operator for the NLS equation has two types of spectra: the discrete (or point) spectrum, which occurs in C + and corresponds to solitons. The continuous spectrum, which in general includes the whole real line, corresponds to the non-solitonic (or radiation) component of the signal. I(λ) R(λ) 7
53 Eigenspace Associated with λ Associated with each point λ in the spectrum of L is a two-dimensional eigenspace in which eigenvectors v evolve in time according to ( ) jλ q(t, z) v t = Pv = q v. (t, z) jλ Assuming q(t, z) is pulse-like, decaying to zero as t, the P operator approaches a diagonal matrix, and the eigenvectors approach v(t, λ) (αe jλt, βe jλt ) T, α, β C. 8
54 Canonical Eigenvectors For example, the canonical eigenvectors at t ( ) ( ) e jλt, e jλ t are bounded for λ in the upper (lower) half-complex plane respectively. We can propagate these vectors forward in time according to v t = Pv, let them interact with q(t, z), and measure what is scattered at t = + in the basis ( ) e jλt, ( ) e jλ t The corresponding coefficients a(λ) and b(λ) are called the nonlinear Fourier coefficients. 9
55 Continuous and Discrete Spectra It can be shown that these scattering projections are well defined at all points λ R, the real line continuous spectrum at those points λ C +, the upper-half complex plane, where a(λ) = (necessarily isolated points due to analyticity of a(λ)) discrete spectrum 5
56 Computing the Fourier coefficients ( v (, λ) e ) jλt ( ) jλ q(t) v t = q v (t) jλ v (, λ) v (+, λ) ( ) e jλt ṽ (, λ ) ( ) e jλt Since the nonlinear Fourier coefficients are time independent, simply compute them at t = +, by projecting v (+, λ) and ṽ (+, λ) onto the basis v (+, λ) and ṽ (+, λ) to obtain [ v (+, λ), ṽ (+, λ) ] = [ ṽ (+, λ), v (+, λ) ] S, where S = ( a(λ) b (λ ) ) b(λ) a (λ. ) 5
57 Nonlinear Fourier Transform For the purpose of developing the inverse transform, it is sufficient to work with the ratios ˆq(λ) = b(λ) a(λ) and q(λ j) = b(λ j) a (λ j ). 5
58 Nonlinear Fourier Transform Definition (Nonlinear Fourier transform) Let q(t) be a sufficiently smooth function in L (R). The nonlinear Fourier transform of q(t) with respect to a given Lax operator L consists of the continuous and discrete spectral functions ˆq(λ) : R C and q(λ j ) : C + C where ˆq(λ) = b(λ) a(λ), q(λ j) = b(λ j) a, j =,,, N, (λ j ) in which λ j are the zeros of a(λ). Here, the spectral coefficients a(λ) and b(λ) are given by a(λ) = lim v t e jλt, b(λ) = lim v t e jλt 53
59 The (ordinary) Fourier Transform The nonlinear Fourier transform measures, for each fixed eigenvalue λ, the interaction of a propagating eigenvector (initialized as a complex exponential) with a waveform q(t). The ordinary Fourier transform measures, for each fixed frequency ω, the interaction of the complex exponential e jωt with a waveform q(t) via the inner product Q(ω) = q(t), e jωt = q(t)e jωt dt. An eigenvector evolution version of the ordinary Fourier transform is recovered by defining ( ) q(t) v t = v, jω giving v = α ( ) + β ( t ) q(τ)e jωτ dτ e jωt 5
60 Example: NFT of a Rectangular Pulse Consider the rectangular pulse { A, t [t, t ]; q(t) =, otherwise. Let T = t t and T = t + t, and let = λ + A. After some work, we find that the zeros of a(λ) in C +, satisfy j tan(t A + λ ) = + A λ, giving rise to the discrete spectrum. The continuous spectrum is given by ( ˆq(λ) = A jλ e jλt ) jλ cot( T ) A Te t +t sinc(tf ). 55
61 Discrete & Continuous Spectra of Rectangular Pulse ˆq(λ) R(λ) I(λ) ˆq(λ) R(λ) I(λ) λ λ A = A = ˆq(λ) R(λ) I(λ) λ A = 6 56
62 NFT Properties Let q(t) ( q(λ), q(λ k )) be a nonlinear Fourier transform pair. (The ordinary Fourier transform as limit of the nonlinear Fourier transform): If q L, there is no discrete spectrum and q(λ) Q(λ), where Q(λ) is the ordinary (linear) Fourier transform of q (t) Q(λ) = q (t)e jλt dt. (Weak nonlinearity): If a, then âq(λ) a q(λ) and ãq(λ k ) a q(λ k ). In general, however, âq(λ) a q(λ) and ãq(λ k ) a q(λ k ). (Constant phase change): ejφ q(t)(λ) = e jφ q(t)(λ) and e jφ q(t)(λ k ) = e jφ q(t)(λ k ). 57
63 Fourier Transform Properties (cont d) (Time dilation): q( t a ) = a q(aλ) and q( t a ) = a q(aλ k ); (Time shift): q(t t ) e jλt ( q(λ, ) q(λ k )); (Frequency shift): q(t)e jωt ( q(λ ω), q(λ k ω)); (Parseval identity): q(t) dt = Ê + Ẽ, where Ê = π log ( + ˆq(λ) ) dλ, Ẽ = N I (λ j ). The quantities Ê and Ẽ represent the energy contained in the continuous and discrete spectra, respectively. j= 58
64 The Key NFT Property (Lax convolution): If q (t) = q (t) (L, M; L), then q (λ) = c(λ, L) q (λ) and q (λ k ) = c(λ, L) q (λ k ). For the NLS equation, c(λ, L) = exp( jλ L). 59
65 Signals and Systems LTI systems: y(t) = h(t) x(t) + z(t) Y (ω) = H(ω)X (ω) + Z(ω) Nonlinear Integrable systems: y(t) = x(t) (L, M; L)+z(t) {Ŷ (λ) = H(λ) ˆX (λ) + Ẑ(λ), Ỹ (λ j ) = H(λ j ) X (λ j ) + Z(λ j ) where the channel filter is H(λ) = e jλz. We see that the operation of the Lax convolution in the nonlinear Fourier domain is described by a simple multiplicative (diagonal) operator (i.e., filters), much in the same way that the ordinary Fourier transform maps x(t) y(t) to X (ω) Y (ω). 6
66 Launching a Pulse (revisited) z = km 8 q(t) t (soliton units) 6
67 Launching a Pulse (revisited) z = km 8 q(t) t (soliton units) 6
68 Launching a Pulse (revisited) z = 5 km 8 q(t) t (soliton units) 6
69 Launching a Pulse (revisited) z = km 8 q(t) t (soliton units) 6
70 Launching a Pulse (revisited) 6 5 z = km Im(λ) Re(λ) q(λ) λ 6
71 Launching a Pulse (revisited) 6 5 z = km Im(λ) Re(λ) q(λ) λ 6
72 Launching a Pulse (revisited) 6 5 z = 5 km Im(λ) Re(λ) q(λ) λ 6
73 Launching a Pulse (revisited) 6 5 z = km Im(λ) Re(λ) q(λ) λ 6
74 Launching a Pulse (revisited) q(t) z = km Im(λ) Re(λ) 3 3 t (soliton units) 63
75 Launching a Pulse (revisited) q(t) z = km Im(λ) Re(λ) 3 3 t (soliton units) 63
76 Launching a Pulse (revisited) q(t) z = 5 km Im(λ) Re(λ) 3 3 t (soliton units) 63
77 Launching a Pulse (revisited) q(t) z = km Im(λ) Re(λ) 3 3 t (soliton units) 63
78 Launching a Pulse (revisited) q(t) z = km Im(λ) Re(λ) 3 3 t (soliton units) 63
79 Nonlinear FDM Immediate application: encode data in the nonlinear spectrum! {λ j, q(λ j ), ˆq(λ)} TX S/P. INFT. P/S D/A LPF channel {λ j, q(λ j ), ˆq(λ)} RX. P/S. NFT. S/P A/D LPF See also: A. Hasegawa and T. Nyu, Eigenvalue Communication, J. of Lightwave Technology, vol., pp , March
80 Multisoliton Communication Multisolitons are signals whose NFT spectrum only contains the discrete component, i.e., the corresponding L operator has only discrete eigenvalues Im(λ) 3 Im(λ).5 q Re(λ) q Re(λ) t 5 5 t sech(t) Satsuma-Yajima -soliton 65
81 Generation of Multisoliton Signals Inverse NFT problem: Given a set of points in C +, compute a signal q(t) such that the eigenvalues of its corresponding L operator are exactly these points Several approaches: Riemann Hilbert system of linear equations, Hirota bilinearization scheme, Darboux transformation Darboux method (recursive construction) Starting with a multisoliton signal q(t, λ, λ,..., λ k ), and a solution, φ(t, λ k+ ), of ( ) jλ q(t, z) v t (t, λ) = q v(t, λ), (t, z) jλ compute q = q j(λ φ k+ λ k+ ) φ φ + φ. q has eigenvalues of q and new eigenvalue from φ 66
82 On-Off Eigenvalue Encoder (OOK) 5 6 Im(λ) 3 b = 3 b = b = b = q(t) Re(λ) 5 5 t 5 6 Im(λ) 3 b = 3 b = b = b = q(t) Re(λ) 5 5 t OOK signals for input bits [ ] and [ ] 67
83 On-Off Eigenvalue Encoder (OOK) 5 6 Im(λ) 3 b = 3 b = b = b = q(t) Re(λ) 5 5 t 5 6 Im(λ) 3 b = 3 b = b = b = q(t) Re(λ) 5 5 t OOK signals for input bits [ ] and [ ] Problem: Spectral efficiency drops as the number of points is increased 67
84 Multi-eigenvalue position encoding 68 Multi-eigenvalue position encoding Im(λ) q(t) t 6 q(t) t.5.5 Re(λ)
85 Multi-eigenvalue position encoding Generate ( N ) + ( N ) + + ( N k ) signals for a given grid of N points, and at most k eigenvalues in a transmit signal Select cut-off bandwidth and pulsewidth SE (in bits/s/hz) 3 3 B c (in GHz) T c (in ns) 3 Spectral efficiency as a function of the cutoff parameters (N=5, =.j, k = 5) 69
86 Multi-eigenvalue position encoding Achievable spectral efficiencies k M M T c B c SE (in bits/s/hz) 3, ns 6.5 GHz.3 5, ns 5.8 GHz.7 5,369,936,569 ns 3.6 GHz 3. Table : Spectral efficiencies of the multi-eigenvalue position encoder (N = 5 and =.) 7
87 Experimental Demonstration ECL laser AWG I Q I/Q modulator EDFA AOM LO Coherent Receiver I x Q x I y Q y Real-time Oscilloscope & offline DSP PC OBPF 3dB AOM Coupler Fiber recirculating loop EDFA 5 km SSMF Raman pump Z. Dong, S. Hari, T. Gui, K. Zhong, M. I. Yousefi, C. Lu, P.-K. A. Wai, F. R. Kschischang, and A. P.-T. Lau, Nonlinear Frequency Division Multiplexed Transmissions based on NFT, IEEE Photonics Techn. Letters, vol. 7, no. 5, pp. 6 63, Aug. 5. 7
88 eigenvalues (3km) (,,,) (,,,) (,,,) (,,,) (,,,) 5 3 (,,,) 5 3 (,,,) (,,,) (,,,) (,,,) (,,,) (,,,) (,,,) (,,,) (,,,) 7
89 3 eigenvalues (8km) (,,) (,,) (,,) (,,) (,,) (,,) (,,) 73
90 Received Signal (8km) 7
91 Recent Result (Yousefi and Yangzhang, 6) Achievable rate [bits/d] 8 6 AWGN NFDM WDM P [dbm] ( km, normal-dispersion, NFDM vs 5 channel WDM in 6 GHz bandwidth) M. I. Yousefi and X. Yangzhang, Linear and Nonlinear Frequency-Division Multiplexing, arxiv:63.389v, May 6. 75
92 Conclusion Pulse train transmission and WDM may not be the best transmission strategy for the nonlinear optical channel. All deterministic distortions are zero for all users under NFDM: NFDM removes inter-channel interference (cross-talk) between users of a network sharing the same fiber channel; NFDM removes inter-symbol interference (ISI) (intra-channel interactions) for each user; with NFDM, information in each channel of interest can be conveniently read anywhere in a network without knowledge of the distance or any information about other users; spectral invariants are remarkably stable features of the NLS flow. Exploiting the integrability of the NLS equation, NFDM modulates non-interacting degrees of freedom and thereby does not suffer from cross-talk, a major limiting factor for prior work. 76
93 Conclusions FPU Zabusky-Kruskal Gardner et al. Lax RM Miura Hirota Zakharov-Shabat Hasegawa Ablowitz et al. WDM, optical solitons Hasegawa s 993 s The NFT can be used for transmission of information NFT solves at least two major problems it removes ISI spectral data appear to be more robust to noise compared with time-domain processing (e.g., digital backpropagation) Is it better than existing methods? We do not know yet. So far we can match existing methods. However, our insights suggest that the achievable rates for NFDM will trend upwards for increasing SNRs. disadvantages: critically relies on integrability it is hard to implement or sometimes to simulate. WDM disp/nonlin man NFT 77
94 Open Problems Plenty of open problems: Entropy (rate) evolution (production) in NLS equation with additive noise Rigorous bounds on capacity Vector models (for polarization, multimode, multicore) Discrete models Development of fast algorithms 78
95 References A. Hasegawa and T. Nyu, Eigenvalue Communication, J. of Lightwave Technology, vol., pp , March 993. R.-J. Essiambre, G. Kramer, P. J. Winzer, G. J. Foschini, B. Goebel, Capacity limits of optical fiber networks, J. Lightw. Technol., vol. 8, pp. 66 7, Sept./Oct.. 3 M. I. Yousefi and F. R. Kschischang, Information Transmission using the Nonlinear Fourier Transform, Part I: Mathematical Tools, Part II: Numerical Methods, Part III: Discrete Spectrum Modulation, IEEE Trans. on Inform. Theory, July. E. Meron, M. Feder and M. Shtaif, On the achievable communication rates of generalized soliton transmission systems, arxiv:7.,. 5 P. Kazakopoulos and A. L. Moustakas, Transmission of information via the non-linear Schrödinger equation: The random Gaussian input case, ArXiv:.79, 3. 6 S. Wahls and H. V. Poor, Fast numerical nonlinear Fourier transforms, to appear in IT Transactions, 5. 79
96 References (cont d) 7 H. Bülow, Experimental assessment of nonlinear Fourier transform based detection under fiber nonlinearity, ECOC. 8 J. E. Prilepsky, et al., Nonlinear inverse synthesis and eigenvalue division multiplexing in optical fiber channels, PRL, Jul.. 9 H. Terauchi and A. Manuta, Eigenvalue modulated optical transmission system based on digital coherent technology, OECC, 3. H. Terauchi, Y. Matsuda, A. Toyota, and A. Maruta, Noise tolerance of eigenvalue modulated optical transmission system based on digital coherent technology, OECC/ACOFT,. S. Wahls, S. T. Le, J. E. Prilepsky, H. V. Poor, S. K. Turitsyn, Digital backpropagation in the nonlinear Fourier domain, SPAWC, Stockholm, Sweden, Jun. 5. V. Aref, H. Bülow, K. Schuh, W. Idler, Experimental demonstration of nonlinear frequency division multiplexed transmission, ECOC 5, Sept. 5. 8
97 6 Steele Prize For their discovery in the 97s of the mathematical framework underlying the nonlinear Fourier transform, Clifford S. Gardner, John M. Greene, Martin D. Kruskal and Robert M. Miura received the prestigious 6 Leroy P. Steele Prize for a Seminal Contribution to Research, awarded by the American Mathematical Society. From the Steele Prize Announcement: Nonlinearity has undergone a revolution: from a nuisance to be eliminated, to a new tool to be exploited. 8
98 Acknowledgment This work was supported: by the Natural Sciences and Engineering Research Council of Canada through the Discovery Grants Program, and by the Institute for Advanced Study, Technische Universität München, through a Hans Fischer Senior Fellowship. I am profoundly grateful for all support. 8
99 Acknowledgment This work was supported: by the Natural Sciences and Engineering Research Council of Canada through the Discovery Grants Program, and by the Institute for Advanced Study, Technische Universität München, through a Hans Fischer Senior Fellowship. I am profoundly grateful for all support. Thank You! 8
Digital Communication over Optical Fibers
Digital Communication over Optical Fibers Frank R. Kschischang Department of Electrical & Computer Engineering University of Toronto frank@ece.utoronto.ca 016 European School of Information Theory Chalmers
More informationStaircase Codes. Error-correction for High-Speed Fiber-Optic Channels
Staircase Codes Error-correction for High-Speed Fiber-Optic Channels Frank R. Kschischang Dept. of Electrical & Computer Engineering University of Toronto Talk at Delft University of Technology, Delft,
More informationSpectral function modulation based on nonlinear frequency division multiplexing
www.nature.com/scientificreports Received: 2 May 207 Accepted: 3 June 207 Published: xx xx xxxx OPEN Spectral function modulation based on nonlinear frequency division multiplexing Guangqiang He,2, Luning
More informationIEEE TRANSACTIONS ON INFORMATION THEORY 1. Information Transmission using the Nonlinear Fourier Transform, Part III: Spectrum Modulation
IEEE TRANSACTIONS ON INFORMATION THEORY 1 Information Transmission using the Nonlinear Fourier Transform, Part III: Spectrum Modulation Mansoor I. Yousefi and Frank R. Kschischang, Fellow, IEEE Abstract
More informationInformation Theory for Dispersion-Free Fiber Channels with Distributed Amplification
Information Theory for Dispersion-Free Fiber Channels with Distributed Amplification Gerhard Kramer Department of Electrical and Computer Engineering Technical University of Munich CLEO-PR, OECC & PGC
More informationarxiv: v2 [cs.it] 21 Sep 2017
Signaling on the Continuous Spectrum of Nonlinear Optical Fiber IMAN TAVAKKOLNIA * AND MAJID SAFARI Institute for Digital Communications, School of Engineering, University of Edinburgh, EH9 3JL, Edinburgh,
More informationSelf-Phase Modulation in Optical Fiber Communications: Good or Bad?
1/100 Self-Phase Modulation in Optical Fiber Communications: Good or Bad? Govind P. Agrawal Institute of Optics University of Rochester Rochester, NY 14627 c 2007 G. P. Agrawal Outline Historical Introduction
More informationarxiv: v1 [cs.it] 26 Jul 2017
Polarization-Division Multiplexing Based on the Nonlinear Fourier Transform JAN-WILLEM GOOSSENS, 1,2,* MANSOOR I. YOUSEFI, 2 YVES JAOUËN 2 AND HARTMUT HAFERMANN 1 1 Mathematical and Algorithmic Sciences
More informationNonlinear Effects in Optical Fiber. Dr. Mohammad Faisal Assistant Professor Dept. of EEE, BUET
Nonlinear Effects in Optical Fiber Dr. Mohammad Faisal Assistant Professor Dept. of EEE, BUET Fiber Nonlinearities The response of any dielectric material to the light becomes nonlinear for intense electromagnetic
More informationFolded digital backward propagation for dispersion-managed fiber-optic transmission
Folded digital backward propagation for dispersion-managed fiber-optic transmission Likai Zhu 1, and Guifang Li 1,3 1 CREOL, The College of Optics and Photonics, University of Central Florida, 4000 Central
More informationOptical solitons and its applications
Physics 568 (Nonlinear optics) 04/30/007 Final report Optical solitons and its applications 04/30/007 1 1 Introduction to optical soliton. (temporal soliton) The optical pulses which propagate in the lossless
More informationMultiuser Capacity Analysis of WDM in Nonlinear Fiber Optics
Multiuser Capacity Analysis of WDM in Nonlinear Fiber Optics Mohammad H. Taghavi N., George C. Papen, and Paul H. Siegel Dept. of ECE, UCSD, La Jolla, CA 92093 Email: {mtaghavi, gpapen, psiegel}@ucsd.edu
More informationNonlinear effects in optical fibers - v1. Miguel A. Muriel UPM-ETSIT-MUIT-CFOP
Nonlinear effects in optical fibers - v1 Miguel A. Muriel UPM-ETSIT-MUIT-CFOP Miguel A. Muriel-015/10-1 Nonlinear effects in optical fibers 1) Introduction ) Causes 3) Parameters 4) Fundamental processes
More informationPropagation losses in optical fibers
Chapter Dielectric Waveguides and Optical Fibers 1-Fev-017 Propagation losses in optical fibers Charles Kao, Nobel Laureate (009) Courtesy of the Chinese University of Hong Kong S.O. Kasap, Optoelectronics
More informationEdinburgh Research Explorer
Edinburgh Research Explorer Capacity Analysis of Signaling on the Continuous Spectrum of Nonlinear Optical Fibers Citation for published version: Tavakkolnia, I & Safari, M 7, 'Capacity Analysis of Signaling
More informationNonlinear Fiber Optics and its Applications in Optical Signal Processing
1/44 Nonlinear Fiber Optics and its Applications in Optical Signal Processing Govind P. Agrawal Institute of Optics University of Rochester Rochester, NY 14627 c 2007 G. P. Agrawal Outline Introduction
More informationNONLINEARITY in optical fibers limits the achievable
5056 JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 35, NO. 23, DECEMBER 1, 2017 Nonlinear Compensation in Optical Communications Systems With Normal Dispersion Fibers Using the Nonlinear Fourier Transform Ivan
More informationJOURNAL OF LIGHTWAVE TECHNOLOGY 1. Impact of Perturbations on Nonlinear Frequency-Division Multiplexing
JOURNAL OF LIGHTWAVE TECHNOLOGY Impact of Perturbations on Nonlinear Frequency-Division Multiplexing Xianhe Yangzhang, Domaniç Lavery, Member, IEEE, Polina Bayvel, Fellow, IEEE, and Mansoor I. Yousefi
More informationOPTICAL COMMUNICATIONS S
OPTICAL COMMUNICATIONS S-108.3110 1 Course program 1. Introduction and Optical Fibers 2. Nonlinear Effects in Optical Fibers 3. Fiber-Optic Components I 4. Transmitters and Receivers 5. Fiber-Optic Measurements
More informationB 2 P 2, which implies that g B should be
Enhanced Summary of G.P. Agrawal Nonlinear Fiber Optics (3rd ed) Chapter 9 on SBS Stimulated Brillouin scattering is a nonlinear three-wave interaction between a forward-going laser pump beam P, a forward-going
More informationSquare Root Raised Cosine Filter
Wireless Information Transmission System Lab. Square Root Raised Cosine Filter Institute of Communications Engineering National Sun Yat-sen University Introduction We consider the problem of signal design
More informationOptical Fiber Signal Degradation
Optical Fiber Signal Degradation Effects Pulse Spreading Dispersion (Distortion) Causes the optical pulses to broaden as they travel along a fiber Overlap between neighboring pulses creates errors Resulting
More informationLecture 4 Fiber Optical Communication Lecture 4, Slide 1
ecture 4 Dispersion in single-mode fibers Material dispersion Waveguide dispersion imitations from dispersion Propagation equations Gaussian pulse broadening Bit-rate limitations Fiber losses Fiber Optical
More informationSupplementary Figure 1: Scheme of the RFT. (a) At first, we separate two quadratures of the field (denoted by and ); (b) then, each quadrature
Supplementary Figure 1: Scheme of the RFT. (a At first, we separate two quadratures of the field (denoted by and ; (b then, each quadrature undergoes a nonlinear transformation, which results in the sine
More informationNonlinear Impairment Compensation using Backpropagation
Chapter Number Nonlinear Impairment Compensation using Backpropagation Ezra Ip 1, Joseph M. Kahn 1 NEC Labs America, Princeton, NJ, USA Stanford University, Stanford, CA, USA 1. Introduction The ultimate
More informationEstimation of the Capacity of Multipath Infrared Channels
Estimation of the Capacity of Multipath Infrared Channels Jeffrey B. Carruthers Department of Electrical and Computer Engineering Boston University jbc@bu.edu Sachin Padma Department of Electrical and
More informationGain dependence of measured spectra in coherent Brillouin optical time-domain analysis sensors
Gain dependence of measured spectra in coherent Brillouin optical time-domain analysis sensors Jon Mariñelarena, Javier Urricelqui, Alayn Loayssa Universidad Pública de Navarra, Campus Arrosadía s/n, 316,
More informationAnalysis of Signal Propagation in Fiber Optic Links. and Compensation Techniques for Propagation. Impairments
Analysis of Signal Propagation in Fiber Optic Links and Compensation Techniques for Propagation Impairments ANALYSIS OF SIGNAL PROPAGATION IN FIBER OPTIC LINKS AND COMPENSATION TECHNIQUES FOR PROPAGATION
More informationEffect of Nonlinearity on PMD Compensation in a Single-Channel 10-Gb/s NRZ System
The Open Optics Journal, 28, 2, 53-6 53 Open Access Effect of Nonlinearity on PMD Compensation in a Single-Channel -Gb/s NRZ System John Cameron *,,2, Xiaoyi Bao and Liang Chen Physics Department, University
More informationRaman Amplification for Telecom Optical Networks. Dominique Bayart Alcatel Lucent Bell Labs France, Research Center of Villarceaux
Raman Amplification for Telecom Optical Networks Dominique Bayart Alcatel Lucent Bell Labs France, Research Center of Villarceaux Training day www.brighter.eu project Cork, June 20th 2008 Outline of the
More informationAccurate Calculation of Bit Error Rates in Optical Fiber Communications Systems
Accurate Calculation of Bit Error Rates in Optical Fiber Communications Systems presented by Curtis R. Menyuk 1 Contributors Ronald Holzlöhner Ivan T. Lima, Jr. Amitkumar Mahadevan Brian S. Marks Joel
More informationCode-Aided Expectation-Maximization and Probabilistic Constellation Shaping for Fiber-Optic Communication Systems. Chunpo Pan
Code-Aided Expectation-Maximization and Probabilistic Constellation Shaping for Fiber-Optic Communication Systems by Chunpo Pan A thesis submitted in conformity with the requirements for the degree of
More informationDual-polarization nonlinear Fourier transform-based optical communication system
Downloaded from orbit.dtu.dk on: Dec 6, 28 Dual-polarization nonlinear Fourier transform-based optical communication system Gaiarin, Simone; Perego, A. M.; Porto da Silva, Edson; Da Ros, Francesco; Zibar,
More informationNonlinearity compensation using dispersionfolded digital backward propagation
Nonlinearity compensation using dispersionfolded digital backward propagation Likai Zhu 1,* and Guifang Li 1,2 1 CREOL, The College of Optics and Photonics, University of Central Florida, 4000 Central
More informationPolarization division multiplexing system quality in the presence of polarization effects
Opt Quant Electron (2009) 41:997 1006 DOI 10.1007/s11082-010-9412-0 Polarization division multiplexing system quality in the presence of polarization effects Krzysztof Perlicki Received: 6 January 2010
More informationSoliton Molecules. Fedor Mitschke Universität Rostock, Institut für Physik. Benasque, October
Soliton Soliton Molecules Molecules and and Optical Optical Rogue Rogue Waves Waves Benasque, October 2014 Fedor Mitschke Universität Rostock, Institut für Physik fedor.mitschke@uni-rostock.de Part II
More informationFiber Gratings p. 1 Basic Concepts p. 1 Bragg Diffraction p. 2 Photosensitivity p. 3 Fabrication Techniques p. 4 Single-Beam Internal Technique p.
Preface p. xiii Fiber Gratings p. 1 Basic Concepts p. 1 Bragg Diffraction p. 2 Photosensitivity p. 3 Fabrication Techniques p. 4 Single-Beam Internal Technique p. 4 Dual-Beam Holographic Technique p. 5
More informationImpact of Dispersion Fluctuations on 40-Gb/s Dispersion-Managed Lightwave Systems
990 JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 21, NO. 4, APRIL 2003 Impact of Dispersion Fluctuations on 40-Gb/s Dispersion-Managed Lightwave Systems Ekaterina Poutrina, Student Member, IEEE, Student Member,
More informationINFLUENCE OF EVEN ORDER DISPERSION ON SOLITON TRANSMISSION QUALITY WITH COHERENT INTERFERENCE
Progress In Electromagnetics Research B, Vol. 3, 63 72, 2008 INFLUENCE OF EVEN ORDER DISPERSION ON SOLITON TRANSMISSION QUALITY WITH COHERENT INTERFERENCE A. Panajotovic and D. Milovic Faculty of Electronic
More informationDigital Band-pass Modulation PROF. MICHAEL TSAI 2011/11/10
Digital Band-pass Modulation PROF. MICHAEL TSAI 211/11/1 Band-pass Signal Representation a t g t General form: 2πf c t + φ t g t = a t cos 2πf c t + φ t Envelope Phase Envelope is always non-negative,
More informationPROCEEDINGS OF SPIE. Optimal input signal distribution and capacity for nondispersive nonlinear optical fiber channel at large signal to noise ratio
PROCEEDINGS OF SPIE SPIEDigitalLibrary.org/conference-proceedings-of-spie Optimal input signal distribution and capacity for nondispersive nonlinear ical fiber channel at large signal to noise ratio I.
More informationCommunication Theory Summary of Important Definitions and Results
Signal and system theory Convolution of signals x(t) h(t) = y(t): Fourier Transform: Communication Theory Summary of Important Definitions and Results X(ω) = X(ω) = y(t) = X(ω) = j x(t) e jωt dt, 0 Properties
More informationProject Number: IST Project Title: ATLAS Deliverable Type: PU*
Project Number: IST-1999-1066 Project Title: ATLAS Deliverable Type: PU* Deliverable Number: D11 Contractual Date of Delivery to the CEC: June 30, 001 Actual Date of Delivery to the CEC: July 4, 001 Title
More informationRevision of Lecture 4
Revision of Lecture 4 We have completed studying digital sources from information theory viewpoint We have learnt all fundamental principles for source coding, provided by information theory Practical
More informationAnalysis and Compensation of Nonlinear. Impairments in Fiber-Optic Communication. Systems
Analysis and Compensation of Nonlinear Impairments in Fiber-Optic Communication Systems ANALYSIS AND COMPENSATION OF NONLINEAR IMPAIRMENTS IN FIBER-OPTIC COMMUNICATION SYSTEMS BY XIAOJUN LIANG, M.A.Sc.
More informationChapter 5. Transmission System Engineering. Design the physical layer Allocate power margin for each impairment Make trade-off
Chapter 5 Transmission System Engineering Design the physical layer Allocate power margin for each impairment Make trade-off 1 5.1 System Model Only digital systems are considered Using NRZ codes BER is
More informationNonlinearity Equalization Techniques for DML- Transmission Impairments
Nonlinearity Equalization Techniques for DML- Transmission Impairments Johannes von Hoyningen-Huene jhh@tf.uni-kiel.de Christian-Albrechts-Universität zu Kiel Workshop on Optical Communication Systems
More informationDifferent Optical Fiber Nonlinear Coefficient Experimental Measurements
Different Optical Fiber Nonlinear Coefficient Experimental Measurements Sugan Shakya, Andis Supe, Ingrida Lavrinovica, Sandis Spolitis and Jurgis Porins Institute of Telecommunications, Riga Technical
More informationDepartment of Electrical and Computer Systems Engineering
Department of Electrical and Computer Systems Engineering Technical Report MECSE-3-4 MODELLING OF POLRISTION MODE DISPERSION EFFECT SYSTEM USING FINITE IMPULSE RESPONSE DICRETE FILTE TECHNIQUE LN Binh
More informationRaman-Induced Timing Jitter in Dispersion-Managed Optical Communication Systems
632 IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 8, NO. 3, MAY/JUNE 2002 Raman-Induced Timing Jitter in Dispersion-Managed Optical Communication Systems Jayanthi Santhanam and Govind P.
More informationChannel Capacity And Modeling Of Optical Fiber Communications
Channel Capacity And Modeling Of Optical Fiber Communications Salam Elahmadi, Mandyam D. Srinath, Dinesh Rajan Department of Electrical Engineering, Southern Methodist University, Dallas, TX 75275 Richard
More informationLECTURE 16 AND 17. Digital signaling on frequency selective fading channels. Notes Prepared by: Abhishek Sood
ECE559:WIRELESS COMMUNICATION TECHNOLOGIES LECTURE 16 AND 17 Digital signaling on frequency selective fading channels 1 OUTLINE Notes Prepared by: Abhishek Sood In section 2 we discuss the receiver design
More informationPrinciples of Communications Lecture 8: Baseband Communication Systems. Chih-Wei Liu 劉志尉 National Chiao Tung University
Principles of Communications Lecture 8: Baseband Communication Systems Chih-Wei Liu 劉志尉 National Chiao Tung University cwliu@twins.ee.nctu.edu.tw Outlines Introduction Line codes Effects of filtering Pulse
More informationNonlinear Photonics with Optical Waveguides
1/44 Nonlinear Photonics with Optical Waveguides Govind P. Agrawal The Institute of Optics University of Rochester Rochester, New York, USA c 2015 G. P. Agrawal Outline Introduction Planar and Cylindrical
More informationPerformance Analysis of FWM Efficiency and Schrödinger Equation Solution
Performance Analysis of FWM Efficiency and Schrödinger Equation Solution S Sugumaran 1, Rohit Bhura 2, Ujjwal Sagar 3,P Arulmozhivarman 4 # School of Electronics Engineering, VIT University, Vellore Tamil
More informationNonlinear effects compensation for long-haul superchannel transmission system. Abdelkerim AMARI
Nonlinear effects compensation for long-haul superchannel transmission system Abdelkerim AMARI April 21, 2016 Abstract Optical communication systems have evolved since their deployment to meet the growing
More informationADAPTIVE EQUALIZATION AT MULTI-GHZ DATARATES
ADAPTIVE EQUALIZATION AT MULTI-GHZ DATARATES Department of Electrical Engineering Indian Institute of Technology, Madras 1st February 2007 Outline Introduction. Approaches to electronic mitigation - ADC
More informationObservation of spectral enhancement in a soliton fiber laser with fiber Bragg grating
Observation of spectral enhancement in a soliton fiber laser with fiber Bragg grating L. M. Zhao 1*, C. Lu 1, H. Y. Tam 2, D. Y. Tang 3, L. Xia 3, and P. Shum 3 1 Department of Electronic and Information
More informationthat efficiently utilizes the total available channel bandwidth W.
Signal Design for Band-Limited Channels Wireless Information Transmission System Lab. Institute of Communications Engineering g National Sun Yat-sen University Introduction We consider the problem of signal
More informationECEN689: Special Topics in High-Speed Links Circuits and Systems Spring 2012
ECEN689: pecial Topics in High-peed Links Circuits and ystems pring 01 Lecture 3: Time-Domain Reflectometry & -Parameter Channel Models am Palermo Analog & Mixed-ignal Center Texas A&M University Announcements
More informationUltra-short pulse propagation in dispersion-managed birefringent optical fiber
Chapter 3 Ultra-short pulse propagation in dispersion-managed birefringent optical fiber 3.1 Introduction This chapter deals with the real world physical systems, where the inhomogeneous parameters of
More informationVector dark domain wall solitons in a fiber ring laser
Vector dark domain wall solitons in a fiber ring laser H. Zhang, D. Y. Tang*, L. M. Zhao and R. J. Knize 1 School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore 639798
More informationEE4512 Analog and Digital Communications Chapter 4. Chapter 4 Receiver Design
Chapter 4 Receiver Design Chapter 4 Receiver Design Probability of Bit Error Pages 124-149 149 Probability of Bit Error The low pass filtered and sampled PAM signal results in an expression for the probability
More informationSystem optimization of a long-range Brillouin-loss-based distributed fiber sensor
System optimization of a long-range Brillouin-loss-based distributed fiber sensor Yongkang Dong, 1,2 Liang Chen, 1 and Xiaoyi Bao 1, * 1 Fiber Optics Group, Department of Physics, University of Ottawa,
More informationA Family of Nyquist Filters Based on Generalized Raised-Cosine Spectra
Proc. Biennial Symp. Commun. (Kingston, Ont.), pp. 3-35, June 99 A Family of Nyquist Filters Based on Generalized Raised-Cosine Spectra Nader Sheikholeslami Peter Kabal Department of Electrical Engineering
More informationDifferential Brillouin gain for improving the temperature accuracy and spatial resolution in a long-distance distributed fiber sensor
Differential Brillouin gain for improving the temperature accuracy and spatial resolution in a long-distance distributed fiber sensor Yongkang Dong, Xiaoyi Bao,* and Wenhai Li Fiber Optical Group, Department
More informationarxiv:quant-ph/ v1 5 Aug 2004
1 Generation of polarization entangled photon pairs and violation of Bell s inequality using spontaneous four-wave mixing in fiber loop Hiroki Takesue and Kyo Inoue arxiv:quant-ph/0408032v1 5 Aug 2004
More informationNonlinearity Mitigation in WDM Systems: Models, Strategies, and Achievable Rates
Nonlinearity Mitigation in WDM Systems: Models, Strategies, and Achievable Rates Marco Secondini Senior Member, IEEE, Erik Agrell Fellow, IEEE, Enrico Forestieri Senior Member, IEEE, Domenico Marsella,
More informationS. Blair February 15,
S Blair February 15, 2012 66 32 Laser Diodes A semiconductor laser diode is basically an LED structure with mirrors for optical feedback This feedback causes photons to retrace their path back through
More informationGain-Flattening Filters with Autonomous Temperature Stabilization of EDFA Gain
Gain-Flattening Filters with Autonomous Temperature Stabilization of A Gain by You Mimura *, Kazuyou Mizuno *, Syu Namiki *, Yoshio Tashiro * 2, Yoshihiro Emori * 2, Mariko Miyazawa *, Toshiaki Tsuda *
More informationSimulation of Pulse propagation in optical fibers P. C. T. Munaweera, K.A.I.L. Wijewardena Gamalath
International Letters of Chemistry, Physics and Astronomy Submitted: 6-- ISSN: 99-3843, Vol. 64, pp 59-7 Accepted: 6--5 doi:.85/www.scipress.com/ilcpa.64.59 Online: 6--5 6 SciPress Ltd., Switzerland Simulation
More informationTheory of optical pulse propagation, dispersive and nonlinear effects, pulse compression, solitons in optical fibers
Theory of optical pulse propagation, dispersive and nonlinear effects, pulse compression, solitons in optical fibers General pulse propagation equation Optical pulse propagation just as any other optical
More informationAnalytical design of densely dispersion-managed optical fiber transmission systems with Gaussian and raised cosine return-to-zero Ansätze
Nakkeeran et al. Vol. 1, No. 11/November 004/J. Opt. Soc. Am. B 1901 Analytical design of densely dispersion-managed optical fiber transmission systems with Gaussian and raised cosine return-to-zero Ansätze
More informationModelling and Specifying Dispersive Laser Cavities
Modelling and Specifying Dispersive Laser Cavities C. Sean Bohun (UOIT), Yuri Cher (Toronto), Linda J. Cummings (NJIT), Mehran Ebrahimi (UOIT), Peter Howell (Oxford), Laurent Monasse (CERMICS-ENPC), Judith
More informationImpact of Nonlinearities on Fiber Optic Communications
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Review Impact of Nonlinearities on Fiber Optic Communications Mário Ferreira
More informationOn Wiener Phase Noise Channels at High Signal-to-Noise Ratio
On Wiener Phase Noise Channels at High Signal-to-Noise Ratio Hassan Ghozlan Department of Electrical Engineering University of Southern California Los Angeles, CA 989 USA ghozlan@usc.edu Gerhard Kramer
More informationWavelength switchable flat-top all-fiber comb filter based on a double-loop Mach-Zehnder interferometer
Wavelength switchable flat-top all-fiber comb filter based on a double-loop Mach-Zehnder interferometer Ai-Ping Luo, Zhi-Chao Luo,, Wen-Cheng Xu,, * and Hu Cui Laboratory of Photonic Information Technology,
More informationLecture 9. PMTs and Laser Noise. Lecture 9. Photon Counting. Photomultiplier Tubes (PMTs) Laser Phase Noise. Relative Intensity
s and Laser Phase Phase Density ECE 185 Lasers and Modulators Lab - Spring 2018 1 Detectors Continuous Output Internal Photoelectron Flux Thermal Filtered External Current w(t) Sensor i(t) External System
More informationNonlinear effects and pulse propagation in PCFs
Nonlinear effects and pulse propagation in PCFs --Examples of nonlinear effects in small glass core photonic crystal fibers --Physics of nonlinear effects in fibers --Theoretical framework --Solitons and
More informationDigital Baseband Systems. Reference: Digital Communications John G. Proakis
Digital Baseband Systems Reference: Digital Communications John G. Proais Baseband Pulse Transmission Baseband digital signals - signals whose spectrum extend down to or near zero frequency. Model of the
More informationInterference Focusing for Mitigating Cross-Phase Modulation in a Simplified Optical Fiber Model
Interference Focusing for Mitigating Cross-Phase Modulation in a Simplified Optical Fiber Model Hassan Ghozlan Department of Electrical Engineering University of Southern California Los Angeles, CA 90089
More informationIndex Terms Nonlinear Schrodinger Equation (NLSE), soliton coupling, all optical switching, split step Fourier method (SSFM).
International Journal of Scientific & Engineering Research, Volume 6, Issue 6, June-015 149 Split-Step Fourier Method in Modeling Dual Core Nonlinear Fiber Coupler James Raja.S. Department of ECE Karunya
More informationModeling Propagation in Optical Fiber using Split- Step Wavelet in Linear Media
International Journal of Electronic and Electrical Engineering. ISSN 0974-2174 Volume 3, Number 3 (2010), pp. 119--124 International Research Publication House http://www.irphouse.com Modeling Propagation
More informationThe Poisson Optical Communication Channels: Capacity and Optimal Power Allocation
The oisson Optical Communication : Capacity and Optimal ower Allocation Samah A M Ghanem, Member, IAENG, Munnujahan Ara, Member, IAENG Abstract In this paper, the channel capacity for different models
More informationPMD Compensator and PMD Emulator
by Yu Mimura *, Kazuhiro Ikeda *, Tatsuya Hatano *, Takeshi Takagi *, Sugio Wako * and Hiroshi Matsuura * As a technology for increasing the capacity to meet the growing demand ABSTRACT for communications
More informationOptical Component Characterization: A Linear Systems Approach
Optical Component Characterization: A Linear Systems Approach Authors: Mark Froggatt, Brian Soller, Eric Moore, Matthew Wolfe Email: froggattm@lunatechnologies.com Luna Technologies, 2020 Kraft Drive,
More informationOn nonlinear distortions of highly dispersive optical coherent systems
On nonlinear distortions of highly dispersive optical coherent systems Francesco Vacondio,,* Olivier Rival, Christian Simonneau, Edouard Grellier, Alberto Bononi, 2 Laurence Lorcy, Jean-Christophe Antona,
More informationChapter 5 Frequency Domain Analysis of Systems
Chapter 5 Frequency Domain Analysis of Systems CT, LTI Systems Consider the following CT LTI system: xt () ht () yt () Assumption: the impulse response h(t) is absolutely integrable, i.e., ht ( ) dt< (this
More information7 The Waveform Channel
7 The Waveform Channel The waveform transmitted by the digital demodulator will be corrupted by the channel before it reaches the digital demodulator in the receiver. One important part of the channel
More informationDark Soliton Fiber Laser
Dark Soliton Fiber Laser H. Zhang, D. Y. Tang*, L. M. Zhao, and X. Wu School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore 639798 *: edytang@ntu.edu.sg, corresponding
More informationChapter 5 Frequency Domain Analysis of Systems
Chapter 5 Frequency Domain Analysis of Systems CT, LTI Systems Consider the following CT LTI system: xt () ht () yt () Assumption: the impulse response h(t) is absolutely integrable, i.e., ht ( ) dt< (this
More informationAutocorrelation Function for Dispersion-Free Fiber Channels with Distributed Amplification
Autocorrelation Function for Dispersion-Free Fiber Channels with Distributed Amplification Gerhard Kramer arxiv:75.454v cs.it Jan 8 Abstract Optical fiber signals with high power exhibit spectral broadening
More informationOptimization of the Split-step Fourier Method in. Modeling Optical Fiber Communications Systems
Optimization of the Split-step Fourier Method in Modeling Optical Fiber Communications Systems Oleg V. Sinkin, 1 Ronald Holzlöhner, 1 John Zweck, 1 and Curtis R. Menyuk 1;2 August 7, 2002 1 Department
More informationSlow, Fast, and Backwards Light: Fundamentals and Applications Robert W. Boyd
Slow, Fast, and Backwards Light: Fundamentals and Applications Robert W. Boyd Institute of Optics and Department of Physics and Astronomy University of Rochester www.optics.rochester.edu/~boyd with George
More informationModule II: Part B. Optical Fibers: Dispersion
Module II: Part B Optical Fibers: Dispersion Dispersion We had already seen that that intermodal dispersion can be, eliminated, in principle, using graded-index fibers. We had also seen that single-mode,
More informationSignal Design for Band-Limited Channels
Wireless Information Transmission System Lab. Signal Design for Band-Limited Channels Institute of Communications Engineering National Sun Yat-sen University Introduction We consider the problem of signal
More informationCROSS-PHASE modulation (XPM) is a nonlinear phenomenon
JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 22, NO. 4, APRIL 2004 977 Effects of Polarization-Mode Dispersion on Cross-Phase Modulation in Dispersion-Managed Wavelength-Division-Multiplexed Systems Q. Lin and
More informationFIBER OPTICS. Prof. R.K. Shevgaonkar. Department of Electrical Engineering. Indian Institute of Technology, Bombay. Lecture: 07
FIBER OPTICS Prof. R.K. Shevgaonkar Department of Electrical Engineering Indian Institute of Technology, Bombay Lecture: 07 Analysis of Wave-Model of Light Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of
More informationAn Efficient Method to Simulate the Pulse Propagation and Switching Effects of a Fiber Bragg Grating
An Efficient Method to Simulate the Pulse Propagation and Switching Effects of a Fiber ragg Grating F. Emami, Member IAENG, A. H. Jafari, M. Hatami, and A. R. Keshavarz Abstract In this paper we investigated
More informationENSC327 Communications Systems 2: Fourier Representations. Jie Liang School of Engineering Science Simon Fraser University
ENSC327 Communications Systems 2: Fourier Representations Jie Liang School of Engineering Science Simon Fraser University 1 Outline Chap 2.1 2.5: Signal Classifications Fourier Transform Dirac Delta Function
More information