Receiver and Amplifier. Dr. Mohammad Faisal Assistant Professor Dept. of EEE, BUET

Transcription

1 Receiver and Amplifier Dr. Mohammad Faisal Assistant Professor Dept. of EEE, BUET

2 Basic Concept: Optical Receiver Optical Receiver converts the optical signal back into electrical form and retrieve the data/intelligence signal transmitted through fiber-optic transmission system. The main element is Photodetector The basic mechanism behind the photodetection is Optical Absorption. The required characteristics of Photodetector: High sensitivity Fast response Low noise High reliability (high efficiency with required BW) Compact size And low cost etc.

3 Photodetector Responsivity If light energy hν>e g an EHP is generated by absorbing the photon. Under influence of an Electric field applying with a voltage, electrons and holes are swept across the semiconductor causing a flow of electric current. This is photocurrent which is directly proportional to the incident optical power P in. Here, R is responsivity of I photodetector (A/W) p RPin

4 Quantum Efficiency: electron generation rate photon incidence rate Ip q h R Pin h q q μm R ; where c h 1.24 R of photodetector increases with λ (as more photons are present for the same optical power) However, with increase of λ, photon energy decreases gradually, and at a certain point photon energy becomes too small to generate electrons, hν<e g then η drops to zero If the facets of the semiconductor slab are assumed to have an antireflection coating, the power transmitted through the slab of width W Where α is absorption coefficient W Ptr e Pin

5 The absorbed power, W P P P 1 e P abs in tr in Since each absorbed photon creates an EHP, the quantum efficiency is Pabs W 1 e P in PhotoDetector A p-n junction under reverse bias is commonly used in practice as PD. The current produced is known as photocurrent.

6 ss PhotoDetector

7 R increases with λ. As λ increases, photon energy decreases, at a certain wavelength photon energy becomes less than the required energy to excite electrons from VB to CB. The wavelength at which R drops to zero is called cutoff wavelength λ c No absorption, so η is zero

8 Rise Time & Bandwidth BW depends on speed with which PD responds to variations of incident optical power Rise time, T r, the time over which the current builds up from 10% to 90% of its final value when incident power is changed rapidly T r is the time taken by electrons and holes to travel to the electrical contacts It also depends on response time of electrical circuit to process the photocurrent Rise Time of PD: T K Constant K depends on RC circuit r tr RC τ tr is transit time, it is the time required for the carriers to be collected after their generation through absorption of photons and τ RC is RC time constant of equivalent RC ckt The BW of PD: f 1 2 tr RC

9 The Transit Time, tr W v d Where W is the width of depletion region, v d is drift velocity. Both W and v d can be optimized to minimize τ tr. W- by controlling concentration of impurities v d by applied voltage The RC Time constant, RC RL Rs Cp R L is external load resistance, R s is internal series resistance and C p is the parasitic capacitance. By proper design these values can be controlled. pn photodiodes can operate up to 40 Gb/s and more Another limiting factor for PD s BW is diffusion current as Diffusion current is related to the absorption of light outside the depletion region Diffusion is inherently a slow process. The carriers generated outside the W region should be collected which takes longer time than that of drift component This part can distort the temporal response of PD

10 Diffusion component can be reduced by reducing the widths of p and n regions and increasing the depletion region so that most of the light is absorbed inside the depletion region and the drift component is enhanced

11 p-i-n Photodiode The depletion region can be increased by inserting a layer of undoped or lightly doped semiconductor material between the p-n junction. Since the middle layer consists of intrinsic material, this PD structure is called p-i-n photodiode

12 The main difference between p-n PD and p-i-n PD is the drift component of photocurrent which dominates over diffusion component in p-i-n structure as most of the incident optical power is absorbed inside the i-region (depletion region). How to control W? Optimum value of W depends on a compromise between speed and sensitivity R can be increased by increasing W, so η can be increased Pabs W q 1 e R ; where c P h 1.24 in But τ tr will be increased with W Consequently BW will decrease. For indirect band gap material (Si, Ge), W~20-50 μm, BW limited by long τ tr > 200 ps For direct band gap material (InGaAs), W~3-5 μm, BW enhanced by small τ tr ~ 10 ps tr W v d

13 The performance of p-i-n PD can be improved by using a double heterostructure design The semiconductor material for middle i-layer is different from p- and n-layers, and its band gap is chosen such that light is absorbed only in the i-layer. As shown in Fig, InGaAs is as middle i-layer and InP for p- and n-layers. InP is transparent for light whose λ>0.92 μm E g for InP is 1.35 ev, E g for lattice matched In 1-x Ga x As with x=0.47 is about 0.75 ev which corresponds to a cutoff wavelength 1.65 μm. This middle layer absorbs strongly in λ of 1.3~1.6 μm The diffusion component is eliminated completely in this structure simply because the photons are absorbed only inside the depletion region InGaAs PDs are very useful in lightwave systems

14 The limitation of p-i-n Photodiode: The BW of p-i-n PD can be increased by decreasing W, but R decreases as η decreases. or R can be increased but at the cost of BW How to solve the problem?

15 Avalanche Photodiode (APD) APD is designed to achieve much larger η and hence R maintaining larger BW PDs with larger R are preferred as they require less optical power Maximum value of R attainable from p-i-n PD is R=q/hν, as η = 1. APD can provide much larger R than that. How? By providing very high electric field inside the device Internal current gain which is known as impact ionization is achieved I p will be increased

16 Primary EHPs are generated by absorption Also called multiplication layer. Secondary EHPs generated through impact ionization E-field in a narrow region

17 Impact Ionization: under high electric field, an accelerated electron produced by photon absorption can acquire sufficient energy to generate a new EHP. The energetic electron gives a part of its kinetic energy to another electron in the VB that ends up to CB, leaving behind a hole. The net result of impact ionization is that a single primary electron, generated through absorption of a photon, creates many secondary electrons and holes.

18 Responsivity of APD R of APD is enhanced by multiplication factor M where M is given as 1 k M A e 1kA ed k A q RAPD MR M h ie d h ka, M i 0 d is the thickness of gain region. α e is impact ionization coefficient of electrons and α h for holes. M= 10 ~ 500, or even more e e

19 Separate Absorption and Multiplication (SAM) region APD Hetero structure design

20 InGaAs material undergoes tunneling breakdown at high E- field ( V/cm which is below the threshold for avalanche multiplication), so InP layer is used for gain region which can sustain quite high (> V/cm) E-field. The absorption region, i-type InGaAs layer, and the multiplication region, n-type InP layer are separate in such a device.

21 Si Reach through APD (RAPD) For Si RAPD, without avalanche gain, Quantum efficiency of nearly 100% in the working region (at wavelength μm)

22 Si reach through APD (RAPD) RAPD consists of p + -π-p-n + layers High electric field region where avalanche multiplication takes place is relatively narrow and centered on the p-n + junction Under low reverse bias, most of the voltage is dropped across the p-n + junction When reverse bias voltage is increased, depletion region widens across p layer and reaches through to the π region (nearly intrinsic/lightly doped) Electric field in the π region is much lower as this region is much wider This causes the removal of some of the excess applied voltage from the multiplication region to the π region giving a relatively slow increase in multiplication factor with applied voltage This is done to reduce noise

23 Limiting Factors Speed limiting factors: Transit time of carriers across the absorption region Time taken by the carriers to cause avalanche multiplication RC time constant In APD, τ tr increases because of generation and collection of secondary EHPs. So BW decreases. f 1, e 2 M e 0 RC τ e is effective transit time and M 0 is low frequency gain. By reducing M 0 BW can be increased Trade off between M and BW Noise is more than that of p-i-n PD

24 APDs advantages +ve/-ve of APD have internal gain, provide much higher sensitivity Wider dynamic range because of their gain variation with response time and reverse bias APDs disadvantages Fabrication is difficult and costly Excess noise due to carrier multiplication Higher bias voltage Variation of M with temperature

25 APD vs p-i-n PD Fig: Minimum detectable optical power for direct detection against transmitted bit rate in order to maintain a BER of 10-9 The figure compares Si PD operating at a wavelength of 0.82 μm where APD is able to approach within 10 to 13 db of quantum limit. p-i-n PD has a sensitivity around 15 db below this level In case of InGaAs, APD requires around 20 db more power than quantum limit, whereas p-i-n PD receiver is some 10 to 12 db less sensitive than APD

26 Problem A PD has a quantum efficiency of 65% when photon of energy J is incident upon it. (a)at what wavelength the PD is operating? (b)calculate the incident optical power to obtain a photocurrent of 2.5 μa at this operating wavelength (c)if the bandgap energy is 0.8 ev at 300 K, determine the wavelength above which PD will cease to operate? The following measurements were taken for an APD, calculate the multiplication factor for the device. Received optical power at 1.35 μm = 0.2 μw Corresponding output photocurrent = 4.9 μa (After avalanche gain) Quantum efficiency at 1.35 μm = 40%

27 Receiver Noise Optical receiver converts optical power into electric current though a PD Receiver produces photocurrent (which incurs the information) along with different noise Two Fundamental Noise mechanism: i. Shot noise or quantum noise ii. Thermal noise Another one: Dark current noise (not major) These noise cause fluctuations of output current produced by PD (The relation I p =RP in assumes no noise) If P in is constant, then shot noise and thermal noise cause fluctuations If P in is fluctuating due to Amplifier noise, then extra noise created at the receiver in addition to those mentioned

28 Shot Noise (Quantum noise) Shot Noise is a manifestation of the fact that an electric current consists of electrons that are generated at random times Then the PD current p s I t I i t Where I p =RP in is the average current and i s (t) is a current fluctuation related to shot noise Detection of light by a photodiode is a discrete process since creation of an EHP results from the absorption of a photon (The energy of quantum of light or photon E=hν). And the signal emerging from the detector is dictated by the statistics of photon arrivals So i s (t) is a stationary random signal with Poisson Distribution, the noise variance is i 2 2 s s 2qI t p f Δf is electrical BW of the receiver

29 Dark Current Noise When there is no optical power incident on the photodiode, a small reverse leakage current flows from the device terminals The dark current also generates a shot noise This is the dark current which also causes random fluctuations around the average flow of photocurrent p d I t I i t Where I p =RP in is the average current and i d (t) is a dark current fluctuation The noise variance is i 2 2 d d 2qI t d f

30 Thermal Noise Thermal noise, shot noise and dark current noise are nearly white and each of them follows more or less Gaussian distribution. Random fluctuations of electrons in any conductir due to thermal agitation Random thermal motion of electrons in a resistor manifests as a fluctuating current (even at the absence of applied voltage) The load resistor in the front-end of an optical fiber receiver adds such fluctuations to the current generated by photodiode. The total current in the receiver, I t I i t i t i t p s d T Where, i T (t) is a current fluctuation induced by thermal excitation. It is a random signal having approximate Gaussian statistics (nearly white) Noise Variance: i t B T T k Tf R Where, k B is Boltzmann constant, T is absolute temperature, R L is load resistor L

31 At the receiver, electrical amplifier is used which adds extra noise. Amplifier noise can be accounted by introducing a factor F n an modifying the equations follows: The total current fluctuations k Tf F 2 4 B n T RL Where, F n is referred to as Amplifier Noise Figure. Physically F n represents the factor by which thermal noise is enhanced by various resistors used in pre- and main amplifiers. I I t I i t i t i t p s d T The total variance of current fluctuations kBTf Fn I 2q I I f 2 s d T p d R L

32 p-i-n Receiver The performance of a receiver is determined by SNR SNR average signal power noise power I p RPin 2 2q RP I f in d 4kBTF R Where R is Responsivity of p-i-n photodiode. Thermal Noise Limit: Assume, σ 2 T>> σ 2 s Thermal noise dominates in most practical cases. Neglecting shot noise term, 2 2 RPR in L SNR, 4k Tf F SNR P B 2 in n 2 2 T s SNR can be improved by increasing the load resistance. That s why most receiver use high impedance front end. L n f

33 Shot Noise Limit: Assume, σ 2 s>> σ 2 T I d is neglected. Let, the pulse energy RPin Pin SNR 2 qf 2 h f E P B N h, as P N hb p in p in p Where 1/B is bit duration, B is bit rate, N p is number of photons contained in 1 bit. Now choosing Δf = B/2, a typical BW for receiver In case of thermal noise limit, RP N phb in SNR N 2 qf 2h B 2 SNR q RLN pb 2k TF N p requires much more in thermal noise limit than shot noise limit for the same SNR B n p

34 SNR I For APD Receiver p M R Pin A in d qm F RP I f I MRP R P p in APD in 4kBTF R L n f s d A in d 2 T 2qM F RP I f where, F k M 1 k 2 1 M 4kBTF R A A A L n f F A is excess noise factor of APD Thermal noise remains same as p-i-n receivers as it originates from the electrical components Shot noise (also dark noise) is enhanced by a factor M as it is generated by photons at random times. Primary EHPs generation by absorption of photons and secondary EHPs generation by impact ionization. Both contribute to the Shot noise, as both cases EHPs are generated at random times

35 Thermal Noise Limit M R Pin RL SNR, 4k TF f SNR P B 2 in n 2 2 T s SNR can be improved by increasing the load resistance. Compared to p-i-n receivers, in APD, SNR is improved a factor of M 2 Shot Noise Limit in 2 2 SNR, and is negligible 2 s T I D 2qM FARPin f RPin = 2 qf f A P 2h F M R P in A f minimum value of F A is 2 and F A increases with M, F A = 1 for noise free case. SNR is reduced by F A compared to p-i-n receivers.

36 Receiver Sensitivity Receiver sensitivity is defined as the minimum average received power P r required by the receiver to operate at a BER of A receiver is said to be more sensitive if it achieves the same performance compared to others with less optical power incident on it. A BER of 10-9 corresponds to on average 1 error per 10 9 bits. For Digital communication: BER exp Q 2 1 erfcq 2 Q 2 where Q parameter or Q-factor is defined as Q I1 I0 Where I 1 is average power for 1 bit, I 0 is average power for 0 bit, σ 1 and σ 0 are the standard deviations of noise for bit 1 and 0 respectively. When Q = 6, BER = 10-9

37 Minimum Received Power Calculate the minimum optical power that a receiver requires to operate at a particular BER reliably I 0 1 0, assume, no optical power at the receiver when bit 0 is transmitted, P 0 0 P1 P0 I1 MRP1 2 MRP, average received power, Pr 2 P is power received when bit 1 is transmitted, M is APD gain RMS noise currents noise current when signal is bit T 2 2 s T noise current when signal is bit s A r 2 4kBTF T R L n f 2qM F R 2 P f ; dark current is ignored

38 Q parameter is given by I1 2MRPr Q s T T 2 2 2MRPr s T T Q Q T Pr qfaqf R M This the analytical expression for P r for a given value of Q for APD For p-i-n receiver: M 1, F 1 P r A Q qqf T R Q T ; if T dominates R 2

39 Sensitivity Degradation The minimum average received optical power (receiver sensitivity) which is derived is based on receiver noise only. However, it may increase due to some non ideal conditions (amplifier noise, optical transmitter s deviations from expected conditions etc.) This increase in average received power is referred to as power penalty. The sources of power penalty which lead to sensitivity degradation even without signal transmission through fiber are: (transmission related power penalty mechanisms are not considered) (1) Extinction ratio (2) Intensity noise (3) Timing jitter Extinction ratio: Power penalty causes due to energy carried by 0 bits. For bit 1, transmitter (say, laser) is ON P 1 (ON-state power) For bit 0, transmitter (say, laser) is OFF P 0 (Off-state power)

40 P 0 P 0 0,for ideal case 0,for practical case, most transmitters emit some power even in OFF-state For semiconductor laser, off-state power depends on bias current I b and threshold current I th. If I b < I th power emitted during 0 bits is due to spontaneous emission, and usually P 0 <<P 1 However, P 0 could be higher if laser is biased close to or above threshold. The extinction ratio is defined as For p-i-n receiver, r ex I RP; I RP ; P r OFF-state power ON-state power P P (for ideal case) P1 P0 2 1 r ex 2RPr The Qparameter is defined as, Q ;(prove it) 1rex 10

41 Consider, thermal noise dominates over other noise so, by using P 1 r 1 r 1 0 The power penalty is defined as r ex ex P p TQ R or ex r r P r P r With the increase of r ex, Pp (db) increases T ex ex 0 0 Pr rex 0 1 r ex 10log 10log db Pr rex 0 1rex Intensity noise We assume that optical power incident on the PD does not fluctuate, but in practice, light emitted by any transmitter shows power fluctuations. Such fluctuations cause intensity noise

42 The PD converts this power fluctuations into current fluctuations which add to those resulting from shot noise and thermal noise As a result, SNR is degraded A simple way to analyze this noise is to add an extra term to the total current variance s d T I where, so, Q I R P 2 in 2RP s d T I T Q is reduced due to intensity noise. So in order to maintain Q for achieving the same value of BER, it is necessary to increase the received power. Timing Jitter So far, we have assumed that the signal (bit sequence) is sampled at the peak of voltage pulse. r

43 In practice, the decision instant for the pulse is determined by the clock-recovery circuit. Due to noisy nature of the input to the clock-recovery circuit, the sampling time fluctuates from bit to bit. This kind of fluctuations is called timing jitter. Sampling time fluctuations cause signal fluctuations, which causes SNR degradation. This can be understood in this way that if the bit is not sampled at the bit center, the sampled value is reduced by an amount that depends on the timing jitter Δt. The SNR can be maintained by increasing the received optical power which increases the power penalty.

44 Problem A silicon p-i-n photodiode incorporated into an optical receiver has a quantum efficiency of 60% when operating at a wavelength of 0.9 μm. The dark current in the device at this operating point is 3 na and the load resistance is 4 kohm. The incident optical power is 200 nw at this wavelength and the post detection filter BW is 5 MHz. Calculate (i) (ii) (iii) (iv) Responsivity of the photodiode Output signal photocurrent rms shot and thermal noise currents at the output SNR at the output when amplifier noise figure is 3 db

45 Optical Amplifier (OA) Amplifier is essential for long distance communication in order to compensate for the fiber losses Previously optoelectronic repeaters (regenerative repeaters) were used in which optical signal is converted into an electric current then amplified/reshaped using electronic amplifier ckt and then converted back to optical signal. (Optical Electrical Optical) It requires optoelectronic devices for Light Source and Light Detector, many electronic circuitry for pulse slicing, retiming and reshaping Such repeaters are quite complex, expensive and it slows down the system. It restricts the operating BW. Optical amplifiers can solve this problem which can directly amplify the optical signal.

46 Optical Amplifier (OA) Basic Concept Most optical amplifiers amplify incident light through stimulated emission (the same mechanism that is used by LASER). An OA is nothing but a laser without feedback The main element is gain medium in which optical gain is realized when amplifier is pumped to achieve population inversion by carrier injection One approach is to use semiconductor laser amplifiers (InGaAsP traveling wave semiconductor laser amplifier TWSLA) Another approach of achieving optical amplification is to use Fiber Amplifier in which gain is provided by either Raman or Brillouin scattering or by rare earth dopants. (Fig. Gain characteristics of OAs)

47 Semiconductor Laser Amplifier (SLA) SLA is based on conventional laser structure where output facet reflections are between 30 and 35%. Different types of SLA Fabry-Perot amplifier, FPA (which is an oscillator biased below oscillation threshold), traveling eave (TW) amplifiers, TWAs which are single pass devices, injection locked laser amplifier (which is a laser oscillator designed to oscillate at the incident signal frequency) FPA is biased below normal lasing threshold current, light enters one facet and appears amplified at the other facet. FPA has resonant nature In TWA SLA, anti reflection coatings may applied to the laser facets to reduce or eliminate the end reflectivities. It works in single-pass amplification mode where FP resonance is suppressed by reducing end facet reflectivity

48 Amplifier Noise All amplifiers degrade SNR of the amplified signal because of spontaneous emission that adds noise to the signal during amplification. Signal Amplifier Signal + ASE Transmission

49 The SNR degradation is quantified through a parameter F n, known as amplifier Noise Figure F n SNR SNR in out Where SNR refers to the electric power generated when optical signal is converted into an electric current. Consider an ideal PD whose performance is limited by shot noise only. Consider an amplifier whose gain is related as P out GPin P in G P out The SNR of input signal is given as s 2 2 I RPin Pin q SNR ; R, 1 2 2q RP f 2h f h in Assume, dark current is zero and f is detector BW (electrical BW)

50 The spectral density of spontaneous-emission-induced noise is (white noise) sp 1 S G n h ν is optical frequency and n sp is spontaneous emission factor n sp N N2 N 2 1 The noise is known as Amplified Spontaneous Emission (ASE) noise which is added as fluctuations to amplified optical signal and then converted to current fluctuations during photodetection process. At the Rx current produced by PD is I R GE E in E sp is a random signal having random phase. Beating of spontaneous signal with signal will produce a noise current I 2R GP E cos in sp sp sp 2 For details, see the text of A. Yariv

51 θ is a rapidly varying random phase. Averaging over the phase and neglecting all other noise sources, the variance of photocurrent is SNR out Amplifier NF 2 4 in sp 2 2 RGP RS f I RGP in GPin 2 4 RGP RS f 4S f P 4S 2 1 spf nsp G in F 2 ; n nsp if G 1 2h f GP G in If n sp =1 (for ideal amplifier, practicaly n sp >1), then F n =2 in sp sp It implies that (SNR) out = ½(SNR) in, i.e., SNR is degraded by 3 db for an ideal amplifier, but practically F n exceeds 3 db and can be 6-8 db. Lower value of F n is expected for lightwave systems

52 Applications of OA Three common applications in OFC systems: As in-line amplifier for long-haul OFC system. As a Power Amplifier or Power Booster to increase the transmitter power by placing an amplifier just after Tx. Transmission length may be 100 km or more depending on amplifier gain and fiber loss As a Preamplifier to receiver to improve the receiver sensitivity. Placing just before the Rx to boost the received power.

53 Applications of OA

54 Erbium-Doped Fiber Amplifier (EDFA) This sort of fiber amplifiers make use of rare-earth elements as a gain medium by doping the fiber core during manufacturing process. Rare-earth elements are Erbium (Er), Holmium (Ho), Neodymium (Nd), Samarium (Sm), Thulium (Tm), Ytterbium (Yb) etc. Erbium is mostly accepted as EDFA s operating wavelength is 1.55 μm Operating Principle:

55 Operating Principle Each level is labeled with corresponding Russel- Saunder s coupling term. When an Er 3+ ion is embedded in an amorphous host material (silica), the individual energy levels are split into a number of sublevels and get broadened to form energy bands. The energy of pump photon is absorbed by the electrons, the electrons raise to higher excited state. After reaching higher energy state, electron rapidly loses part of its energy nonradiatively and fall to metastable level. if pump energy is high, population in the metastable state will exceed that in the ground state- population inversion is achieved.

56 An intense pumping source i.e. a laser emitting 0.98 μm can be used to excite Er 3+ ions from the ground state 4 I 15/2 to pump band 4 I 11/2 (a). The excited ions decay non-radiatively in about 1 μs from pump band to metastable band (b). The life time of ions in this band is very long (about 10ms) 4 I 15/2 to 4 I 13/2 transitions (c) can be achieved using photons of wavelength 1.48 μm. The absorption of this pump photon excites an electron from the bottom of the ground state to the lightly populated top of the metastable band. (e) spontaneous emission may occur, some ions may de-excited in absence of photon and fall back randomly to ground state with emission of photons 1.55 μm. (f) signal photon may trigger an excited ion to drop back to ground state stimulated emission. (it occurs in the wavelength range 1.53 μm ~ 1.56 μm)

57 EDFA Spectrum

58 Pumping Scheme/Configuration: Pumping at a suitable wavelength provides gain through population inversion. The gain spectrum depends on pumping scheme as well as presence of other dopants like germania or alumina within the core Efficient pumping is done using semiconductor lasers operating at and 1.48-μm wavelengths Most EDFAs employ 0.98 μm pump lasers which can provide 100mW power. Efficiency as much as 11dB/mW by μm μm pump lasers are available but requires larger fibers and high pump powers Three Configurations: Forward Pumping: Pump and signal beams propagate in the same direction. Backward Pumping: Pump and signal beams propagate in opposite directions. Bidirectional Pumping: Pumping is done from both directions (forward and backward) and signal beam propagates in one direction only.

59 Raman Amplifier Fiber-based Raman amplifier uses SRS effect in silica fibers under intense pumping If the pump power is above the threshold value, the Raman gain can exceed the losses and the scattered beam gets amplified. The power of the stimulated Raman line has been found to be much greater than that of spontaneous emission. EDFA vs FRA In EDFA, stimulated emission occurs where an incident signal photon stimulates emission of another identical photon without losing its energy In FRA, SRS occurs where the incident pump photon gives up its energy to create another photon of reduced energy at a lower frequency, the remaining energy is absorbed by the medium in the form of molecular vibrations FRA does not require population inversion

60 Operating Principle The pump and the signal beams at wavelengths λp and λs are injected into the fiber through a fiber coupler. The energy is transferred from the pump beam to signal beam through SRS effect.

61 Fiber p s p s As a result the power at λs increases. In other words, if a suitable optical fiber is optically pumped by an appropriate source, the signal beam will get amplified as the two beams co-propagate along the fiber

62 Raman Gain Spectra

63 Raman Gain Spectra DCF is 8 times more efficient than a standard SMF because of its smaller core diameter Frequency dependence of Raman gain is almost same for the three kinds of fibers. The gain peaks at a Stokes shift of about 13.2 THz The gain BW is about 6 THz (FWHM)

64 Large BW of FRA But it requires relatively large pump powers longer fibers can be used in order to reduce required high powers The figure shows an example of gain spectrum measured for a FRA made by pumping 12 diode lasers. Nearly flat gain profile over a huge 80 nm BW

65 Semiconductor Optical Amplifier (SOA) In Laser, the amplifying medium is confined with optical feedback mechanism to create a resonant cavity (FP cavity) in which light passes back and forth and gets amplified, finally coherent light is taken out In SOA, the light signal passes through the amplifying medium only once.

66 Structure of SOA Double-hetero structure (DH) configuration Band gap of active layer is lower than that of confining layers and its refractive index is larger than that of confining layers giving waveguiding structure

67 Under forward bias, carriers (electrons and holes from n-type and p-type materials) travel towards the active layer where they get trapped in a low-band-gap potential well Biasing current is increased enough to create population inversion in the active layer Signal photons passing through the active layer can stimulate radiative recombination of electrons and holes resulting in the amplification of signal power The signal beam must be coupled efficiently into and out of the SOA chip, usually to a single-mode fiber Optical feedback must be suppressed. All measures must be taken to reduce optical reflections at the facets of the active layer to less than 0.01%

68 Gain Spectrum Gain exhibits ripples implying the effects of residual facet reflectivities Broad gain spectrum

69 Some comments about SOA SOAs are polarization sensitive. The amplifier gain G differs for TE and TM modes. This feature is undesirable for lightwave systems in which state of polarization changes with propagation along the fiber The use of SOA as a preamplifier is attractive since it permits monolithic integration of SOA with the receiver. However, the noise figure is relatively large (5-7 db) for such configuration SOAs can be used as power amplifier to boost the transmitter power. However, it is difficult to achieve in excess of 10mW power. SOAs as in-line amplifiers are not attractive because of polarization sensitivity, interchannel crosstalk and large coupling losses As an optical signal processing element, they have many applications, such as, for wavelength conversion, line code conversion, fast switch for wavelength routing in WDM network etc.

70 Lightwave Transmission System Transmission System IM/DD System Coherent System/Phas e Modulated System Modulation Formats ASK (OOK) FSK PSK (DPSK, QPSK, DQPSK, QAM)

71 Most of the currently available optical fiber transmission systems are IM/DD systems which use OOK format Recently phase modulated transmission systems or coherent systems have been studied and tested a lot (although research on coherent systems is not new) OOK is a simple format in which information lies in the amplitude. its transmitter and receiver configurations are straight forward. But the receiver sensitivity is low. Furthermore, OOK-based transmission system is vulnerable to dispersion and nonlinearity, and at high bit rate like 40 Gb/s or more, intrachannel nonlinearities cause severe performance degradation. That's why, OOK data format is not suitable for high speed optical transmission.

72 FSK modulation technique has relatively higher receiver sensitivity but at the expense of complex transceiver configurations. Moreover, bandwidth expands drastically with the increase of number of channels. For these reasons, FSK data format may not be so popular for high speed fiber-optic WDM and dense WDM networks In PSK format, information is coded into the phase of the carrier signal. It has a constant envelope with compact spectrum. Furthermore, it is more robust to dispersion and nonlinearity. PSK with differential scheme enables simple direct detection with increased OSNR. However, it has some drawbacks, like precise alignment of transmitter and receiver which is complex, stringent requirement of laser linewidth etc.

73 IMDD System PSK system is relatively complex. Coherent detection is necessary, however, incoherent detection is possible by utilizing differential scheme.

74 PSK-Based Transmission System OOK modulation format for optical carrier to be modulated by intelligence signal Simple detector at the receiver

75 The phase of the optical carrier signal generated by laser diode is modulated by the digital data information. In binary PSK, when data changes from 1 to 0 or vice versa, the phase of the carrier is altered to 180 degree and thereby information is encoded into the phase. This modulated signal is transmitted through fiber and received at the receiver. For PSK, either homodyne or heterodyne coherent detection is used, which is complex and costly. That s why, differential encoding of the phase modulated signals are performed and preferred as they enable simple direct detection with enhanced OSNR.

76 Multichannel Communication System The huge potential bandwidth of optical fiber can be efficiently utilized by multiplexing a number of channels and transmitting them through the fiber simultaneously.

77 The transmission bandwidth of fiber is divided into a number of nonoverlapping frequency (or wavelength) bands and each of these bands is associated with an optical carrier to support a single communication channel. System capacity can be increased more than 10 Tb/s Classification of Multichannel Transmission Systems: Multichannel System OFDM WDM OTDM OCDM

78 OFDM In FDM the optical channel bandwidth is divided into a number of nonoverlapping frequency bands and each signal is assigned one of these bands of frequencies. The individual signal can be extracted from the combined signal by appropriate electrical filtering or optical filtering at the receiver terminal. Hence FDM is usually done electrically at the transmit terminal prior to intensity modulation of a single optical source. In case of broadcast communication, tunable optical filters or local oscillators (LO) can be used, and for point-to-point communication, fixed optical filters can be used. FDM is usually used for low capacity, short distance broadcast communications. FDM and WDM schemes differ from each other in respect of transmitter/receiver configuration.

79 Optical FDM System Information source 1 Laser 1 f 1 Receiver Information source k.... Information source N Laser k Laser N f k f N M U X Optical fiber D e M U X Frequency selective DeMux (LO-1) Receiver. N. (LO-N)..

80 WDM Point-to-point high capacity WDM link. The role of WDM is to increase the total system capacity (bit rate) The ultimate capacity of WDM system depends on channel spacing and total number of channels The minimum channel spacing is limited by interchannel cross talk

81 WDM Channel 1 Channel 1 Source 1 Detector 1 Channel 2 Source 2 Optical MUX Optical Fiber Optical DeMUX X Detector 2 Channel 2 Channel N Channel N Source N Detector N Block diagram of a unidirectional WDM network Input Channel Output Channel Source 1 Detector 2 WDM Device 1 Optical Fiber WDM Device Detector 1 Source 2 Output Channel Input Channel 2 Block diagram of a bidirectional WDM network

82 WDM Network POP: Point of Presence (Central office), ADM: Add/Drop Multiplexer The links between Nodes consist of fiber pairs or multiple fiber pairs

83 Optical Network Optical Networks are used to connect a large number of users They may include: LAN, MAN, WAN depending on the area to cover. LAN: a few km, MAN: a few hundred Km, WAN: several hundred to thousands km. They can be designed using the hub, ring or star topology, usually for MAN and WAN ring topology, for LAN - star topology Several LANs are connected to a MAN by using passive wavelength routing, several MANs are connected to a WAN whose nodes are interconnected using mesh topology In optical networks, nodes are connected through point-topoint WDM links; optical cross-connect (OXC) switches are used In case of electronic processing, the maximum speed achievable is 10Gb/s, however, all-optical signal processing supports much more than 10Gb/s

84 WDM networks that use passive star coupler are often called passive optical networks (PON) as they avoid active switching. Multiple wavelengths are used for routing the signals in the local loop called Passive Photonic Loop

85 WDM Coarse WDM (CWDM) Δλ>0.4 nm (50GHz) Dense WDM (DWDM) Δλ 0.4 nm (50GHz) Spectral Efficiency: s B f ch Bit Rate Channel spacing Say, bit rate is 10 Gb/s. Spectral efficiency for CWDM, Δf ch =100 GHz (0.8 nm) so η s = 10/100 = 0.1 b/s/hz = 10% DWDM, Δfch=50 GHz (0.8 nm) so η s = 10/50 = 0.2 b/s/hz =20% Larger spectral efficiency is expected.

86 OTDM TDM is usually done in electrical domain. But electrical TDM system becomes difficult to implement at bit rates above 10 Gb/s (because of limitations imposed by high speed electronics) Optical TDM can solve this problem which can increase the bit rate of a single optical carrier to more than 1 Tb/s

87 OTDM Transmitter N channels at a bit rate B of each share the same carrier frequency and are multiplexed optically to form a composite bit stream at the bit rate NB. The laser should capable to generate a periodic pulse train at the repetition rate equal to the single channel bit rate B. The laser should produce pulses of width T p such that T p <T B to ensure that each pulse will fit within its allocated time slot T B. The laser output is split equally into N branches A modulator in each branch blocks the pulses representing 0 bits and creates N independent bit streams at bit rate B Multiplexing of N bit streams is achieved by a delay technique. The bit stream in the nth branch is delayed by an amount (n-1)/(nb), where n=1,2,,n. The output of all branches is then combined to form a composite signal

88 OTDM mux is built with SMF except modulators which is made of LiNbO 3 or semiconductor waveguides Splitting and recombination are done by 1 N FBT coupler Delay lines are implemented using fiber segments of controlled lengths OTDM technique requires the use of RZ format OTDM requires optical sources emitting a train of short pulses at a repetition rate as high as 40 GHz. Gain switching or mode locking of a semiconductor lasers can provide such pulses (10-20 ps) with high repetition rate

89 OTDM Demultiplexing OTDM DeMux: Electro-optic or All-optical techniques All-Optical: OTDM demuxing using FWM as a nonlinear medium The OTDM signal is launched together with CLOCK signal into a nonlinear medium (HNLF). The CLK is a pump signal with different wavelength for FWM process

90 In time slots in which a clock pulse overlaps with the 1 bit of the channel to be demultiplexed, FWM produces a pulse at a new wavelength. As a result, the pulse train at this new wavelength is an exact replica of the channel to be demultiplexed. An optical filter is used to separate the demultiplexed channel from the OTDM and clock signals.

91 OCDM Each channel is coded in such a way that its spectrum spreads over a much wider region than by the original signal. In WDM technique, bandwidth is partitioned to share by users. In TDM, time slots are allotted (dynamic or fixed) to the users. All users share the entire bandwidth and all time slots in a random fashion in case of OCDM

92 Several methods can be used for data coding, such as, direct sequence coding, time hopping and frequency hopping. The figure shows the DS for optical CDM. Spectral spreading is accomplished by means of a unique code which is independent of input signal. The decoder uses the same code for compressing the signal and recovering the data. The spread spectrum code is called signature sequence. Such coded signal is difficult to jam or intercept. That s why CDM signal is secured. The same spectral BW is used by many users distinguished on the basis of different signature sequences assigned to them. The signature sequences come from a family of orthogonal codes so that each signal can be decoded accurately at the receiver. The decoding is accomplished using an optical correlation technique.