Photonics and Optical Communications - Science topic

what is your simulator and simulation criteria?

Supercontinuum generation by short pulse with high power that lead to traction or fusion soliton.

the commercial glass optical fiber is comsist of core and cladding, core is doped with GeO2, cladding is pure solica. the pure silica can be used under 1600C but the dopant can't. If you want to monitor the temperature base on the optical fiber, I suggest you use FBG sensor. After writing FBG on the commercial optical fiber and then you can regenerate the FBG, then the FBG sensor can be used under 1000C. There is another option is to use the PCF which is made from pure silica.

I don't know

Thank you so much Amirhassan Shams-Ansari for answers.

I am encountering few more issues:

Using COMSOL or MATLAB or other simulation softwares it is easy to calculate the MFD. You need consider the change of wave-guiding difference as the tapering diameter decreasing: initially silica/(silica+Ge) and then air/silica. I believe you cannot use a simple formula to get the accurate result

What is confinement loss?

i am also comfuse about the beta valume, how can we calculate the beta2, beta 3...

Dera Mustafa Hammood ,

you may ahve a look at equation 1a and 1b for a description of the gap and wavelength dependence of the coupling loss/ transmission in a butt joint SM-fiber connection:

But, sorry, no experimental data yet...

Best regards

G.M.t

Dear Gerhard Martens ,

After seeing your and other respectable researchers' answers, I am glad I asked this question.

I am really delighted to hear from you the history of an ever-lasting discussion about sensor and actuator definitions. I have always found it annoying that the sensor definition has usually been preferred as a "too specific" definition to serve only for an interface of an electrical/electronic system and an "other" system/medium with different form of signal(s).

Besides, that diiscussion, I can start another one:

There are many commercial integrated devices which are called "sensor"s, although in fact they are not basic sensors but are more complicated small systems which may also include electronic amplifier(s), filter(s), analog-digital-converter, indicators etc. For sure, these are very convenient devices for electronic design, but I think it is not correct to call them "sensor". Such a device employs a basic sensor but besides it provides other supporting electronic stages to aid the electronic designer. I don't know if there is a specific name for such devices.

Thank you again for your additional explanations.

Best regards...

The reliability of data transmission characterizes the probability of getting a distortion for the transmitted data bit. This indicator is often referred to as the Bit Error Rate (BER). The BER value for communication channels without additional means of error protection is 10-4 — 10-6, in optical fiber — 10-9. A ber value of 10-4 indicates that on average, one bit is distorted out of 10,000 bits. The q-factor of the receiving system Q is determined from the expression:

Q = GA/TC,

or, in logarithmic form:

Q[dB] = GA[dB] - 10lgTC[x].

It is the q-factor of the receiving system that determines the signal-to-noise ratio (C/N) at the output of the low-noise Converter (LNC or LNB). It is important to note that the final C/N value does not depend on the LNC gain.

The Mach-Zender modulator can be modulated in different ways. It depends on your task.

For example, I simulated a modulator with a differential delay line having different geometric shoulder lengths. But at the same time, a sinusoidal signal was supplied to the optical input of the modulator, which modeled the generation of a laser diode. If you put a loss cell in one channel of the trim line, which you can change with an external signal, then you get an approximate model of the Mach-Zehnder EOM electro-optical modulator.

In chapter 6 of my book, I describe in detail the Mach-Zehnder EOM modulator.

Look at the link.

"Laser Optoelectronic oscillators"

Alexander Bortsov, Yuri I'lin, Sergei .M. Smolskiy

Springer Nature Switzerland AG 2020

A. A. Bortsov et al., Laser Optoelectronic Oscillators, Springer Series in Optical

Sciences 232, https://doi.org/10.1007/978-3-030-45700-6

And then describe in detail your task.

Alexander Bortsov

My quasi D-Shaped fiber is this according to the paper i follow for its design but i need this kind of result which i attached in the second screen shot. Please, some one help me in troubleshooting the problem?

Hello Avijit Prakash,

Based on my knowledge, the radiation pattern of your light source has a significant effect on your results. Therefore, I strongly agree with Dr. Sascha

Regards. - Hossien

Hello again

To solve the above problem,

First, enable the (''use linear wavelength spacing'') option as shown below, then pay attention to the rest of the description:

Determine the number of frequency points according to the type of divisions of your wavelength range. For example, if you want the distance to be 0.5 nano, then subtract the beginning of the interval from the end of it, then multiply by 2, and add 1 at the end.

Finally, the number of frequency points is determined.

For example, above is the number of frequency points:

(900-280)*2+1=1241

Thanks

Motahari

One may use MS Visio, its a basic tool and the image quality is good.

As an update to the wonderful white paper that Mohammad Albooyeh shared above, there are a few typos in his pdf worth mentioning. In equation 3 the cos(kai*k*L) term in the numerator should be deleted, as should the sin(kai*k*L) term in the numerator of equation 4. Fig 1 should also be updated, and I have attached a corrected version. Note that this is amplitude, not intensity, and S31/S41 are the simulated results. This particular figure was done with a mesh of lambda/10, but I was able to achieve comparable results with a mesh as coarse as lambda/5.

Zhijia Hu ，我不是图片文章的作者，麻烦把我从中剔除掉，谢谢！

what is the approximate speed of the soliotn in fiber optics? OR Does it depends on the fiber nonlinearities ??

Hi Dr. Dasgupta,

Have you solved this problem?

I am wondering how to simulate EM field near the back focal plane (e.g. a few micrometers away from the BK plane )

Dear Anuj, the point raised by you is really timely. From your points, it is understood that traditional methods will not help much. However, I would like to say QAM may be an option along with orbital angular momentum multiplexing. Just have a look at the articles attached.

a) https://www.osapublishing.org/oe/viewmedia.cfm?uri=oe-26-2-2099&seq=0

b)

Depending on which direction you want to observe your E or H fields distribution (phase and magnitude) you can define your reference plane at any distance you want (by modelling a shaped plane along xoy,xoz..) then define your monitor at the frequency of interest and finally run the simulation. After simulation you can explore the results in the navigation tree as usual.

Hope this helps!

Hi Sibasish,

Although the reply from Christoph is technically correct, it does not really answer your initial question.

The reason for your question lies in the fact that most book and articles (even those that are not written for the layman, let alone all the internet scribbles by so-called experts ) seldom stick to clear and standard terminology. Many people do indeed use the terms colour space when then mean colour model and vice versa.

As you might know, colour is not a primary physical property but a perceptual result that is said to have three perceptual dimensions: hue, chroma and lightness. In the digital world, however, colour is represented using global colour models and more specific colour spaces. Colour is often mathematically defined as a three-dimensional (3D) property.

However, a point in a 3D space can be defined using many different coordinate systems (CSs). In cartography, a CS defines a set of axes used (i.e. the dimensionality of the system) and the attributes of these axes (direction, name, abbreviation, units, order). As an example: a three-dimensional Cartesian XYZ system using meter. If any of these attributes changes, the coordinate system changes.

The same goes more or less for colour. As such, many CSs for colour exist, built upon three or four coordinate axes. A specific set of coordinate axes used to describe colour is called a colour model. Although a colour model is thus a particular CS to describe colour (e.g. RGB, CMYK), it only describes its dimensionality (three or four variables/axes) and the specific name, direction and order of its axes. Colour models can thus be thought of as systems that describe how colours can be mathematically represented. Most models use three different axes to represent one colour (like the RGB colour model which uses red, green and blue axes), while others such as CMYK use four components/axes.

However, a specific RGB triplet such as R:50 – G:25 – B:255 does not define a particular colour, but only indicates the ratio of the three components used. Although 255 is the maximum value for each R, G and B channel in an 8-bit image, a colour model does not specify how ‘vibrantly blue’ this maximum should be? In other words: a colour model is a mathematical system with no connection to any reference system of colour so that there is no description at all on how these values should be interpreted. Therefore, colour models such as RGB, CMYK (using the primaries cyan, magenta, yellow and blacK/Key), HSV (hue – saturation – value) and HSL (hue – saturation – lightness) are said to be relative.

To accomplish an unambiguous description of colours, the RGB values need to be defined with respect to a particular scale of reference, more specifically a well-defined colour standard. Generally, the CIE XYZ or CIE L*a*b* colour spaces are used as reference standard since they define all the colours an average human perceives. Thus, as soon as the colour model has an associated mapping function that maps its arbitrary values to absolute CIE XYZ values, a colour space is born. In practice, this means that the units of the axes are explicitly described. Since the exact description of a particular colour is accomplished by numbers that are given in relation to a specific colour space, colour spaces are absolute descriptions.

A colour space is thus much more specific than a colour model, as it is just one possible instance of the more general colour model. Using different mapping functions, various absolute colour spaces can be created. Without going into further detail, the following elements are essential for describing a colour space.

The best-known colour spaces that use the RGB colour model are sRGB, Adobe RGB (1998) and ProPhoto RGB (also known as ROMM RGB). The attached figure compares all three in the CIE (x,y) chromaticity diagram. The chromaticity coordinates (x,y) are defined by

x = X / (X + Y + Z)

y = Y / (X + Y + Z)

so that the chromaticity of a colour can be described irrespective from its luminance. The three points that form one triangle correspond to the chromaticity coordinates of the three primaries of that particular colour space (these values can also be found in the attached figure). The area in the triangle is known as the colour gamut and encompasses all the colour values that can be created by mixing those three primaries. Values outside the triangle are said to be out-of-gamut for these primaries. Despite the criticism on such diagrams, they are still one of the most straightforward – although not very accurate – means to compare different gamuts on a sheet of paper.

From this figure, it can be inferred that a colour space such as sRGB has a rather limited gamut, while Adobe RGB [1998] and ProPhoto RGB have much wider gamuts. In other words: the ProPhoto RGB colour space can represent more colours than the Adobe RGB [1998] space which can in turn store a wider range of colours than the sRGB colour space. Despite these differences, all three are based on the RGB colour model to mathematically represent colours.

I think when nonlinear refraction is dominant, you can extract the nonlinear parameters from the closed z-scan with some confidence. However, there are cases , for example when either NL refraction or absorption are dominant, that you cannot do that without ambiguity, so that is why it is customary to run the open z-scan to get the NL absorption parameters first, and then used them in the closed-aperture results. Experimentally all you need is a beam splitter in the far field, and an extra detector yo obtain both the open and closed-zscan traces at the same time

I really appreciate your help with my project, Prof.Zbigniew Motyka and Prof.Maria Chiara Ubaldi.

Dear Dr. Ali Benghia

I suggest you book Physics of Nonlinear Optics, Guangsheng He, Song H. Liu-World Scientific

Dear Prof. Zbigniew Motyka

I appreciate all of your help. Thank you for your presentation.

Your explanation is so good. However, there is another explanation of the Professor Qusay M Ali Hassan as follows:” A pump laser beam with Gaussian intensity distribution is able to stimulate a phase shift, ∆ϕ, in the shape of bell in a nonlinear medium in the transverse direction with respect to the direction of the beam as the one shown in Fig. 1.

On a Gaussian curve, for any point, ρ1, there exists another point, ρ2, having the same slope and wave-vector, their radiation can interfere. Destructive and constructive interferences occurs when the change in phase from the point ρ1, ∆ϕ(ρ1), and the one from the point ρ2, ∆ϕ(ρ2), when the relation ∆ϕ(ρ1)- ∆ϕ(ρ2)= mπ (where m is a constant being odd or even integer for destructive and constructive interferences respectively) is verified.”

I think the diffraction effect does not affect the radius measurement because diffraction only redistributes energy but does not result in energy loss.

The CW laser is probably inducing a thermal lens in your sample, leading to negative nonlinear refractive index. Short pulse laser excitation with high repetition rate (MHz) may also induce thermal lens. To avoid thermal effect you should use a short pulse laser with low repetition rate (10Hz) or explorer other technic as Thermally managed Z-scan.   

Dear Shivaraj R Maidur,

You also investigate optical limiting with CW laser.

THANK YOU, Dr. RAAD.

This paper for birefringence fiber application..

There should be a unique value for the phase mismatch, but as Michael Reilly points out, you need to include sufficient terms in the Taylor expansion.

The calculation is simplified if the Taylor series is expanded about the mid frequency f_m = (f₁ + f₂) / 2

In this case the odd orders of the propagation constant (group velocity and dispersion slope) cancel, as do the zero order terms.  The second derivative of the propagation constant (chromatic dispersion) will dominate in most cases.  For very widely spaced frequencies close to the zero dispersion wavelength, it may be desirable to include the 4th derivative (dispersion curvature term).

Are you working directly with the Taylor coefficients, or using a published formula including dispersion and dispersion slope?  Some widely cited papers from 30 years ago present formulas with sign errors in the dispersion slope term, and are inaccurate near the dispersion zero wavelength.

Dear Vikram,

I agree with colleague Fabrice, so i suggest you attached files in topics.

Best regards

fortunately i found the answer.

I am attached the answer below.

For my research, the range of frequency is 90 to 100 kHz and operating between 200 to 300 Volts

I'm still in planning either to use directional coupler or not. If it available in market for this system specification, then maybe i will purchase it.

By the way, thanks for your suggestion (high voltage differential probe). Maybe this can be used to estimate maximum voltage in the system.

Typically, the effective refractive index is the effective refractive index of the guided mode.  If you want to learn more about simulating waveguides, take a look at Lectures 11 and 12 here:

I sent you a book and I hope it will be useful to you

Hi!

You can define port BC for each head/tail of PC coupler, then you'll have S-Parameters according to the where you inject EM waves.  Transmission and reflection coefficient can be defined in the terms of these parameters. Here is an example...

Port 1= input and Port 2=Output, transmittance and reflectance can be expressed as,

T=abs(S21)^2 ,   R=abs(S11)^2

Dear Dr. Raad

The optical limiting depend on the RSA or Two photons absorption phenomena , any materials dye , crystal and NPS you can use, if this phenomena generated inside material when interaction with high intensity power.

Regard

Dr. Jassim

Thanks Dr. Raad ,

In short once can say that the PCF relies on symmetry. And this is severely broken by bends. As a consequence the (photonic) band-gap structure changes.

Thank osamah for these recommendations articles,

the first part are intensive presentations for under graduated students as  a level as I have had presentation for my 4th year student.

the second an following papers and reports are useful who wanted to get carried in the Lasers Q-switching and the lasers applications.

it was my pleasures this your contact and suggestions.

let my hear you more and more 

best wishes for scientific carriers 

Raad

please read the ch.9 from this book

Dear Taif,

Many thanks for your reply. I haven't shared the complete code as I am assuming that I am making the mistake somewhere in the split step method that's why I have shared only the lines.

I am working on this paper and trying to replicate the figure 3 (phase shift):

C. Lacava, M. J. Strain, P. Minzioni, I. Cristiani, and M. Sorel, "Integrated nonlinear Mach Zehnder for 40 Gbit/s all-optical switching" (also attached with this message).

Can you please confirm that are my inputs right? i.e.

signal_0 = A0_signal.*sech(t/T0).*exp(-0.5i*chirp_p*(t/T0).^2);

probe_0 = A0_probe.*exp(-0.5*(1+1i*chirp_p).*(t/T0).^(2*mshape));

Thank you

Umair Ahmed

As far as I know, if you are using some metal nanoparticles for SERS, it's the LSPR wavelength that matters. This wavelength decides the enhanced field at that wavelength.

The Raman enhancement factor is the product of gain at excitation and scattering frequency. Please have a look to this: http://aip.scitation.org/doi/10.1063/1.1571979

 

Silicon Photonics Design: From Devices to Systems by Lukas Chrostowski is a good starting point. However, more information on your need would be helpfull to give you advice. 

Dear Andrew Metcalf, Thank you for the detailed answer.You answered precisely what I wanted to understand.Now things are much clear.

For the HE11 mode, group index is equal to core group index at a specific V-number which is a function of the core-clad index difference.  Snyder & Love "Optical Waveguide Theory" (1983) includes approximations valid in the weakly guided limit in their table 14-3, page 313.  If you know Δ, solve for V.  Then calculate distortion parameter using their formula in the same table, and hence find waveguide component of group velocity dispersion using equation 11-58 on page 229.   Table 14-4 or figure 14-3 may be more convenient if you don't need exact values.

More generally, if you know core and clad materials, core diameter and core-clad index difference, solve the wave equation at your operating wavelength.  Calculate group velocity and dispersion - either by differentiating the propagation constant with respect to optical frequency, or by applying the mode field integrals presented by Snyder & Love in table 11-1 or 13-2.

The group index and dispersion of core and cladding materials can be calculated by interpolating between the compositions measured by Fleming "Material dispersion in waveguide glasses", IEE Electronics Letters, vol 14, no 11, pp 326-328 (1978).  Use Fleming's Sellmeier coefficients to calculate refractive index, group index and dispersion at any wavelength.

This is where the answer changes from "have you tried google", to "it depends".

There are two main contributions to dispersion in single mode fibres. 

In a real optical fibre these contributions interact, and other factors such as the wavelength dependence of the relative index difference Δ must be included.

In a step index fibre (as specified in your previous question), and in the absence of material dispersion, waveguide dispersion is a minimum for normalised frequency V₀ approximately 3 - just beyond the theoretical frequency for single mode operation (V_C = 2.4048).  Group velocity is a minimum at V~3, so pulse delay decreases with increasing frequency further into the multimode region (anomalous dispersion), and delay increases with increasing frequency (normal dispersion) in the single mode region.

In the absence of material dispersion, if you increase Δ, and keep everything else constant, the cut-off wavelength will shift to longer wavelengths, and so will the dispersion zero wavelength:  λ₀  = π d sqrt( 2 n₁Δ )  /  V₀  where V₀ ~ 3.

If material dispersion is present, then the result depends on both normalised frequency (V = π d sqrt( 2 n₁Δ ) / λ) and on Δ. 

The magnitude of waveguide component of dispersion is proportional to the the product D V Δ, where D is a dimensionless function of V, and is only weakly dependent on Δ (independent of Δ in the weak guidance limit where  Δ tends to zero).

For a step index fibre operating at V=1.9, an 11% increase in Δ will cause a 5% increase in V, and reduce D by a factor 0.79, for a net reduction in waveguide dispersion by 8%.  If the dispersion zero wavelength is in the single mode region it will be shifted to shorter wavelengths.

In contrast, if Δ is increased, but the core size is reduced to keep V-number (and cut-off wavelength) constant, then the magnitude of waveguide dispersion will increase in proportion to Δ.  Zero dispersion wavelength in the single mode region will be shifted to longer wavelengths.

Note that dispersion depends on the second derivative of the propagation constant, and waveguide dispersion is rather sensitive to details of the refractive index profile.  Diffusion of dopants during fibre manufacture makes it difficult to achieve a perfect step index profile, especially for larger values of core refractive index.

More details including results for graded profile single mode fibres in A. W. Snyder & J. D. Love "Optical Waveguide Theory" (1983)  http://www.springer.com/gb/book/9780412099502

The optical path difference of 1-3 and 1-4-2-3 may be the key.  Eliminating the resonator can also be possible with the path length difference longer than the coherence length.

Dear,

Nikolay, why don't you try Comsol Multiphysics software. In this software, you will find wave optic module as well as radio frequency module which will be useful for simulation

Optical Power meter is simple enough tool, hardly any more complicated than digital multimeter. Typically, it consists of two parts - a meter itself that displays the power, and a remote light sensor connected to the meter with a cable. You can position the sensor anywhere you need. On a meter, you only need to set wavelength you're measuring at.

And you don't have to buy it - I'm pretty sure that ThorLABS or Newport power meters can be rented. Or even borrowed.

I don't think you can simulate OPM or use off-the-shelf PIN photodiode as a substitution. The sensor that comes with any OPM is precalibrated (with each sensor having individual calibration!), that's what makes it a measurement tool and not just light detector. If you don't know in advance spectral response of the photodiode - you can't get reliable results.

Dear Haidar Taki,

I grateful for helping me and the UWB is the part of my project.

I will read this paper hoping to benefit from it

Thank you 

It depends on.

Thanks a lot Haniel..........

If I understand your configuration correctly, your laser is connected to one port of the directional coupler.  A few metres of optical fibre are wound around a piezo-electric cylinder, with ends connected to input and output ports of the directional coupler.  The remaining coupler port is connected to a detector whose electrical output is monitored by an oscilloscope.

How stable is the temperature?  The temperature coefficient of refractive index of silica is around 8.5 10^-6 K^-1 (Leviton, 2006)  The linear thermal expansion coefficient of silica is 0.5 10^-6 K^-1.   In 5 m fibre, the optical path length will change by 45 10^-6 m K^-1 , so roughly 0.3 wavelengths for 0.01 K temperature change.  Temperature stability better than 0.001 K is desirable.

The thermal expansion coefficient of the PZT could further increase the temperature sensitivity.

How stable is the laser temperature and drive current?  A typical 1550 nm DFB might have temperature coefficient of wavelength around 0.1 nm / K (12 GHz / K optical frequency coefficient).  Free spectral range of a 5 m fibre resonator cavity is 0.33 pm, or 0.04 GHz. 

The change of DFB optical frequency with drive current might be 0.6 GHz / mA, so current stability of a few parts per million, and temperature stability better than 0.0001 K is desirable, depending on the phase stability you require.

Note also that stretching the fibre in the PZT will change the fibre birefringence.  If the input polarisation is not aligned with the birefringence axes, the polarisation state of the output will vary with PZT drive.  This will change the magnitude of the demodulated signal.  If normal (not polarisation maintaining) fibre is used, the output will be sensitive to disturbance and temperature change in the fibre between laser and directional coupler.

If you are using COMSOL Multiphysics, it is simple to calculate. Just use RF module and go for modal analysis. Effective index comes directly.

You get the green angle from the positions of the top and bottom of the dish.  The shape of the dish is a parabola, with the equation y squared = 4 f x, where f is the distance from the origin to the centre of the feed.  If you know the height yb at the bottom of the dish you can get xb from that equation.  The same is true for yt and xt at the top of the dish.  The angle is arctan((xt-xb)/(yt-yb).  The feed usually points close to the centre of the reflector, but perhaps a little above that to compensate for the longer distance to the top rim of the reflector compared to the bottom rim, which reduces the power density at the top of the reflector if the feed points directly at the middle.

In general, the higher the refractive index contrast, the wider the photonic band gaps.  The lattice constant does not have an affect on the band gap, but it has everything to do with where the band gaps occur in frequency.  The larger the lattice spacing, the lower the frequency of the band gaps.  Notice that I am using the plural version of band gap because many lattices have multiple band gaps.

For more information on this, take a look at Lecture 14 here:

Before watching the video or looking through the notes, you may want to work through lectures 7, 8, and 12 that cover some preliminary topics.

Hope this helps!

After checking the parameters of transmitter and receiver, you also have to take care of parameters of fiber specially the dispersion. 

Hi Ahmed, It's been a while now. How're you doing? Hope all is good.

You're right in that dispersion makes different frequencies travel at different speeds in a medium. That's called chromatic dispersion (there are other mechanisms, like mode dispersion, and also PMD).

Now, you're talking about optical communications. It's implicit then that we're talking about signals that carry information. How can a signal carry information if its spectrum consists on a single frequency? Let's use a simple example; imagine that you turn on a CW laser, launch it through a long optical fiber and observe the output. What would you expect to see? You'll see a constant level of optical power at a constat frequency. How's that conveying any information to the other side? If you want to send information to the other end of the fiber, you need to modify one or more of the signal's properties in time, i.e., you need to modulate it. The faster you modulate that laser, the faster you can transfer information, and that, in general, requires a larger optical bandwidth. Larger bandwidth means lots of different frequencies traveling through the fiber.

Dispersion can be a bad thing, not only in communications. In general, the shorter an optical pulse, the higher the impact of dispersion during its propagation through a medium. But dispersion can be engineered. One can design the properties of a material and the geometry of a waveguide in order to control its dispersion. And we've been using it to design systems and devices for a very long time. That's also part of the reason why fiber telecomm links are mostly exploited in the 3rd window (around 1.5 um). Standard single-mode optical fibers are most transparent at those wavelengths (that's where the absorption minimum of fused silica is). The zero dispersion wavelength of standard SMF fibers is at 1.3 um, as you correctly point out. But fibers are more lossy in that region. So we use the lower loss window to transmit (which is more critical, because then you have to amplify less), and compensate the dispersion using special fibers, or devices such as linearly-chirped fiber Bragg gratigns. This is all very well-known, in fact, nowadays the problem has shifted to compensate the nonlinear effects that occur in the transmission, due to the ever increasing demand for higher data rates, which tends to require more channels, which in turn means more power in the fiber. When you mix dispersion with nonlinearity you have lots of other effects going on.

Hope this was useful. Cheers!

The best for the modeling of radiation losses induced by bend is to use a mode solver  in cylindrical coordinates with boundary conditions like PML or TBC (TBC are better). Commercial software (Lumerical, rsoft, VFEM of Optiwave).

Please find 2 attached files with some details of the method use in a mode solver.

Regards

What type of AWG do you have, and by how much do you wish to change the centre frequency?

For small frequency shifts, temperature tuning may be an option.  Some AWGs use a Peltier device to stabilise the temperature, offering a means to control the centre wavelength.

Many manufacturers put a great deal of effort into making their AWG devices athermal, in which case temperature tuning is unlikely to be effective. 

If the frequency shift is from one ITU channel frequency to another, then an optical switch to select an appropriate port on the remote AWG is an option.

Alternatively, the AWG can be specifically designed with a large temperature coefficient to allow tuning over a wide frequency range, for example as described by Zhang et. al.

Another approach is to employ a liquid crystal as the tuning mechanism, as reported by NTT, Dai and others.

Horn et. al. fabricated in Lithium Niobate for an electrically tunable device.

Thank you Abdelhalim Zekry for answering my question 

Your idea has sense.

For efficient coupling into single mode fibre, it is important to optimise the overlap between the input beam and the mode field supported by the fibre.  This is often managed by matching the numerical aperture of the beam and fibre mode.  If beam and mode field are both Gaussian in profile, an NA mismatch as large as 20% reduces the theoretical coupling efficiency by only 4% (to 96%).

Different criteria apply for coupling into a large core highly multimode fibre.  The aim here should be to under-fill the fibre, not to match the NA.

If the input beam is Gaussian, and the NA (measured at 1/e² intensity) equals the fibre NA, then 14% of the light is lost.  If the fibre NA is only 0.8 of the beam NA, then 20% of the light is lost. 

For efficient coupling into a step index multimode fibre

For a collimated Gaussian beam, wavelength λ, diameter D at lens with focal length f, the diffraction-limited focused spot diameter is  d = 4 λ/π f/D

For an f/4 lens (NA 0.125) underfilled by a factor 2 (converging beam half-angle 0.0625 radians) the spot size for 1 micron light is 10 microns, so requirement 3 is readily met for 100 micron and larger core diameters.

If the focused spot diameter is too small, higher angle rays will be excited.  Even if these fall within the theoretical numerical aperture of the fibre, high order propagating modes are more susceptible to bending loss than low order modes.

Propagation losses will depend on the wavelength, fibre length and fibre composition.  Flame hydrolysed silica (high-OH) has absorption around 1 dB/m around 950 nm, and significantly higher around 1250 and 1384 nm. 

Low-OH silica has losses less than 0.02 dB/m over a wide spectral range, with higher values at blue and ultra-violet wavelengths.

Expect Fresnel reflection losses of 4-5% at each glass-air interface.  An uncoated single-element lens has two such surfaces, and the entrance and exit facets of the fibre introduce a further 2 surfaces.  Fresnel reflection losses of 8-20% are possible, depending on the refractive index of the lens, and whether it is anti-reflection coated.

Normally the transmission/reflection of BS is not affected by the mode of the beam. Exceptions are inhomogeneous polarization over the beam and large spectral width. Of course if your beam splitter and/or waveplate is not homogeneous you are in trouble. A typically weak effect with the diode lasers is that you have an unpolarized background <<1% typically. So if you use a PBS to suppress nearly all the light this background becomes important.

Dear Penglong Ren,

Since you are using collimator, In fact using collimator has advantages  such as robust assembly,  reduced risk of damage to the fiber and reduced return loss.

I have read this article, and I found it has good details regarding your problem  

Since I don't know what your application, I suggest reading about the FiberLock if you are working on modulation.

Best regards