List of muddy points in optical communications
Contents
Introduction
If you found something confusing in an Opt Comms Eng lecture and thought it was a rather "muddy point" then ask a question on this page. Anyone can answer: either myself or any student or anyone in the department can answer. I will of course tweek everyone's answers to make them clear and check they are correct. So don't worry, just go for it and have a stab at it! You are allowed to correct my answers and make them better.
Wave-particle duality
Question: Why is wave/particle duality such a pain in the butt?
Answer: I totally agree with you! Philosphically it is very painful. However, for optical comms, all you need to remember is that you use the wave model for propagation and the particle (photon) model for detection or emission.
Here is a really really crude picture to help you remember this: think of the ocean—the water flows/propagates as waves—however when the waves crash on the beach (ie. are "detected") the waves break up into a spray of water droplets ("photons"). Now this analogy is not perfect from a physics viewpoint, but is cute enough for this engineering subject.
Couplers
Question: What use is a coupler and what is it?
Answer: Basically it is just a lens that improves the efficiency of optical coupling between the fibre and the light detector or light source.
RFI and EMI
Question: What is RFI and EMI?
Answer:
RFI = Radio Frequency Interference EMI = ElectroMagnetic Interference
The point being that fibre optics is immune to both of these nasty things.
Cable versus fibre
Question: Are fibre optics much more useful than metal cable?
Answer: Yes. You get more bandwidth....so you can put multimedia services thru' fibres. Also, as we mentioned in the lectures, fibres are free of many disadvantages that cables have.
Size
Question: What is the size of optic fibre?
Answer: I assume you mean diameter? See Section 5.8 of the text, as there are various sizes. In Table 5.4 you'll see it ranging from 5 to 1000 microns. You may come across cables that look a lot thicker.....but that is because of the sheath thickness and also you can get multiple fibre cores.
Security
Question: I thought that a fibre is not entirely secure: when it is bent at a sharp angle it emits light.
Answer: Good one. What I was saying is that fibres are more privacy secure than cables, as with cables you can non-invasively detect the signals going thru' them via the external fields. However, you are right, there is a very teeny halo around the fibre: this is called the "evanescent" wave. I doubt information could be tapped from this field, but if you put a sharp bend in the fibre you may get some detectable light.....this may not really be that non-invasive, as you run the likely risk of breaking the fibre. The principle is that if you bend the fibre, the angle of incidence, at the bend, may be past that of the critical angle of total internal reflection. Hence you get some light escaping.
Fibre size notation
Question: What does 125 [math]\mu[/math]m fibre at 0.82 [math]\mu[/math]m mean? (slide 13).
Answer: The 125 [math]\mu[/math]m refers to the diameter and the 0.82 [math]\mu[/math]m refers to the best spectral transmission window for that fibre.
Dispersion
Question: I don't understand dispersion.
Answer: Read Section 3.2 of the textbook. Basically, the point is that the refractive index of a fibre is actually not constant with with spectral wavelength. Now we know that speed of propagation, [math]v=c/n[/math], where [math]n[/math] is the refractive index. So some spectral wavelengths will travel faster than others. This is called dispersion.
Information capacity
Question: Why does spectral width imply limited info capacity?
Answer: Now information capacity is determined by the range of modulation frequencies we can send down the fibre. The highest modulation frequency we can send is when the delay between the slowest and fastest spectral wavelengths is half the modulation period (see Fig. 3.9). With this delay period, we see from the this figure that we get cancellation—thus our information capacity is limited.
Spectrum
Question: What do you mean by spectrum?
Answer: The word "spectrum" literally means "range." So when we talk about "electromagnetic spectrum" we mean the complete range of wavelengths or frequencies from uhf (long wavelength) to cosmic rays (short wavelength). So when we talk about the "optical spectrum" we mean the complete range of wavelengths that we can see. Sometimes you'll hear me talk about "spectral wavelength,"...this refers to the wavelength in the spectrum, which must be distinguished from "modulation wavelength."
I know this is confusing to have two types of wavelengths...but we all have to get used to it! Sometimes in the lectures, you may hear me make a slip-of-tongue and say one when I mean the other...do call out and stop me, if I do this.
BER
Question: What does BER mean?
Answer: BER = Bit error rate. So for instance, if we have a BER of [math]10^{-6}[/math], it means that 1 in every million bits that we detect is in error. As you can have digital transmission thru' optical fibres, we can use BER as a measure of quality. We'll be doing this in more detail in future lectures.
Transmission frequencies
Question: What's the significance of transmission frequency?
Answer: The point here is that only certain spectral frequencies can propagate thru' the fibre. The fibre will look opaque to some spectral frequencies. The trick is to choose the spectral frequencies at which the fibre is most transparent. We call these "transmission windows." If we operate outside a transmission window, we'll get lots of attenuation...and so we avoid that. All materials are more absorbing at certain frequencies than at others...and this all has to with resonances in the material.
Rise time of RZ
Question: Since the allowable pulse spread is 35% of the time slot in RZ is system rise time = 0.35*T ?
Answer: I'm not 100% sure, but I think there are two effects here; one is from pulse spreading, one is from the system rise time. They are independent effects, but they both force you to choose a certain rate for your RZ function, R_rz.
I believe the f_3dB fundamentally is determined from the frequency spectrum of the pulse signal you are sending, and I think it so happens that the formula f_3dB (elec) = 0.35/(t_r) gives the right number, as does 0.35/(delta_tau).
Material Dispersion, M value
Question: Does the value of M given during the exam?
