| Contents:
introduction |
Light wave communication was first considered more than 100 years ago. the implementation of optical communication using lightwave guides was restricted to very short distances prior to 1970. Corn in g Glass company achieved a breakthrough in 1970 by producing a fuse dsilica(SiO) fiber with a loss of approximately 20dB/km. the development of semiconductor light sources also started to mature at about that time, allowing the feasibility of transmission over a few kilometers to be demonstrated. Since 1970, the rate of technological progress has been phenomenal, and optical fibers are now used in trans-oceanic service. Besides the long-distance routes, fibers are used in the inter CO( interexchange) routes, and the subscriber loop is the final link in what will eventually be the global interconnection chain . optical fibers are associated with high-capacity communications. A lot of attention is presently being given to optical fibers to provide a very extensive broadband ISDN.
The evolution of optical fibers has been extremely rapid over the past 20 years. Research and development have been directed toward reducing the signal attenuation of fibers, and also toward increasing the digital transmission rate through the fibers.Until the development of optical amplifiers, the attenuation defined the distance between regenerators, which directly affected the cost of an optical fiber route. Initial costs increase as the number of regenerators increases, so the regenerators pacing must be maximized. Note that in regenerators for optical fiber systems there is an optical -to- electrical conversion, complete regeneration of the pulse stream, retiming, and electrical back to conversion, ready for onward transmission. Maintenance costs also increase as the number of regenerators increases. Optical amplifiers do not entirely eliminate regenerators, but they do radically increase the distance between them. The transmission rate (bits per second) directly determines the number of channels the link can carry, so research has been aimed at maximizing parameter. The bit rate is dependent on the l inch e-width of the lightsource the size and dispersion characteristics of the fiber. Dispersion causes transmitted pulses to spread and over lap as they travel along a fiber, limiting the maximum transmission rate.
Ease of coupling light in to the fiber is also an important factor. This relates to the diameter of the light-carrying portion of the fiber and the characteristic of the glass wave guide.A parameter called the numerical aperture is associated with coupling.Top
Spectral width of the transmitting source.
The optical source does not emitan exact single frequency but is spread over a narrow band of frequencies. The laser has a narrower line width than the LED, and consequently is said to produce less in tramodal or chromatic dispersion than the LED. If the light sources could emit only one frequency, there would be no dispersion problem. Unfortunately, that is not possible.
Characteristics of the fiber.The optical fiber causes dispersion, and the term chromatic dispersion is used to relate the spectral width of the source to the properties of the fiber. Chromatic dispersion is defined as the extent to which light pulses are spread for a specific source line width in an optical fiber due to the different group velocities of the different wave lengths within the source spectrum. the total chromatic dispersion is the sum of three components:
1.Modal dispersion
2.Material dispersion
3.Wave guide dispersion
Modal dispersion is dependent only on the fiber dimensions or, specifically, the core diameter.Single-mode fibers do not have modal dispersion. Multi-mode fibers suffer modal dispersion because each mode travels a different distance along a fiber and therefore has a different propagation time. This was the first type of fiber to be produced. Because the refractive index in the core is constant, the velocity of each mode is the same, so as the distance traveled by each mode differs from one to another, so does the propagation time. Because the light in these larger core fibers is composed of several hundred different modes, a pulse becomes broader as it travels along the fiber. The graded index profile causes the light rays toward the edge of the core to travel faster than those toward the center of the core. This effectively equalizes the transit times of the different modes, so they arrive at the receiver almost in phase. The graded index fiber pulse broadening (i.e.dispersion) is significantly improved over the step- index design.For multimode fiber, modal dispersion is a major limitation to high bit-rate performance. The single-mode fiber when operating in a truly single mode, has no other interfering rays, so it does not suffer modal dispersion pulse broadening. The single-mode fiber does, however, suffer from material and wave guide dispersion. This is because of the frequency dependence of the refractive index (and therefore the speed of light) for the fiber material. For silica, the total dispersion drops to zero at 1.31um.TopMulti-mode fiber has a larger core diameter, which is the part of the wave guide carrying most of the signal strength( the remnant being carried in the cladding). This means that light signals entering one end of the fiber can reach the destination in a variety of ways: straight down the wave guide, though a series of reflections at the cladding interface, or in corkscrew fashion. The result is dispersion of the signal, spreading of the optical pulse in time. This limits the data rate or liquid band width.
Single-mode fiber, in contrast, has a core diameter (under10µm) which is small enough to prevent this & multi path effect. Sin gle-mode fiber is the type used in long haul telecommunications, such as in trans-oceanic telephone cables.
But is the tremendous band width of single-mode fiber really needed in underwater applications? The deepest point in the ocean is still a short haul from the surface in fiber optic terms. The answer to this question lies in the selection of the telemetry systems components, the most important one being the cable.
Single-mode vs. Multi-mode Fiber
It would appear that multi-mode fiber has an adequate bandwidth - distance produced to serve even the most demanding(Sub-sea and Land -based) requirements. A system having a 10,000 meter tether length can operate at 40MHz (or40Mbps) and still satisfy the 40MHz-km criteria.
The other key factor, however is attenuation. Single-mode fiber has lower attenuation then multi-mode fiber, as seen in table 1. But even here, the attenuation (signal loss) in multi-mode fiber is acceptable for many deep ocean systems, where in there is typically a flux budget of 18 or20 dB with standard low cost components. Why, then has single-mode fiber been chosen for some deep applications? The answer lies in its reduced susceptibility to micro bending losses.Micro bending
Micro bending loss is the increase attenuation that occurs when the fiber is bent. For every fiber a critical angle within which the light will be reflected back with in the core and retained but beyond which it will be refracted out in to the cladding. It can be seen that local bending can increase this loss by increasing the angle. Single-mode fiber is intrinsically less prone to such losses than multi-mode as the light may be considered to be travelling purely axially.
The type of bending which most marine cables experience when wound over sheaves or outer drums does not in crease the loss significantly. Micro-bending loss is a smaller scale phenomenon reflecting the design of the cable. The optical fiber may be subjected to micro-bending when it is pressed against other cable elements due to external pressure. This effect, while very small, is cumulative over the cable length, and generally increases with pressure. A multi-mode fiber cable, to operate well in deep applications, must be designed to minimize micro-bending loss by providing for each fiber an environment that is free from localized bending. This can be achieved, at least in part, by the design of the plastic insulation surrounding the fibers, the choice of strength member (Kevlar -vs- steel) and its construction (braid -vs- parallel lay), the material between the fibers and strength member and other factors. The cable manufacturers can give advice and recommend designswhich have proven themselves.Cable-Handling
The designer of the under sea telemetry system must be confident that all of the essential components making up the system are available and proven. These include cables, connectors/couplers, transmitters/receivers and cable-handling equipment. An essential part of the
handling gear is a slipring or rotary joint in the winch to permit continuous operation of the under water system regardless of how much cable has been deployed.
Conclusion
The trade off between multi-mode and single-mode fiber optic telemetry is fairly complex. Multi-mode has the advantages of easier interconnection due to the larger core, lower cost components and generally adequate specifications for most under sea applications. Single-mode has the advantage of even greater bandwidth and range specifications, reduced attenuation and perhaps most importantly less micro-bending loss under pressure. Designers should work with suppliers of all sub-system components, including the rotary joints. The building blocks are in place to allow the adoption within confidence of either.
FiberOpticEquipmentComponents
As stated earlier, the two light sources available are the semi-conductor LD and the LED. Both devices have small physical dimensions, which make them suitable for optical fiber transmission. As the term diode suggests, the LDs and LEDs are pn junctions. Instead of being made from doped single crystals, they now have exotic combinations of two or more single-crystal semi-conductor materials. These hetero junctions are consequently called hetero-structures. The fundamental difference between an LD and an LED is the fact that the light from an LED is produced by spontaneous emission, where as light from an LD is made by stimulated emission. This results in the laser having an out put that is coherent and therefore has a very narrow spectrum, where as an LED has an incoherent output and a wide spectrum. The selection of an LD or LED for an optical transmission system depends upon the following factors:
|
Required output power |
|
|
Coupling efficiency |
|
|
Spectral width |
|
|
Type of modulation |
|
|
Linearity requirements |
|
|
Band width |
|
|
Cost |
The semi-conductor LED can be used in the surface-emitting or edge-emitting mode depending upon the type of fabrication. The surface-emitting style has good temperature stability and low cost. However, the coupling efficiency into the fiber is limited by its wide active area. The light power coupled from a commercially available LED into a single-mode fiber is about 100uW. The light power output is incoherent (i.e., the output is spread over a wide spectrum of about 40nm). The operational bitrate is limited by the parasitic capacitance with in the LED.
The edge-emitting LED has improved performance compared to the surface-emitting type. The structure can achieve a higher coupling efficiency into a single-mode fiber, and the narrower active layer compared to the surface-emitting style has a smaller capacitance, which allows higher bitrate operation. The low cost and improved temperature characteristics of the edge-emitting LED compared to the LD have stimulated a lot of research to improve the devices so that they can be used for fiber- in - the -loop(FITL).
For an LED to achieve are a sonable transmission distance (more than 10km between regenerators) at high bitrates (more than 622Mb/s), single-mode fiber, operating at the zero dispersion wave length, must be used. Multi-mode fiber operation significantly reduces the bitrate-regenerator distance product.
Structure and emission modes of al ight emitting diode.
LaserDiode.Top
The LD has evolved extremely quickly over the past decade. The development of the LD is central to the present-day long-distance capability of optical telecommunications. When the current density within the active region of the diode reachs a certain level, the optical gain exceeds the channel losses and the light emission changes from spontaneous to stimulated (i.e.,lasing). the threshold current which this occurs is quite low in the double hetero-structure semiconductor lasers and is typically 5to15mA at25°C. This is a very undesirable characteristic because it means that the drive current must be increased as the temperature increases in order to maintain a constant output power. The internal power dissipation within the diode itself contributes to an increase in temperature, so a runaway situation can occur if some form of temperature control is not used. Also, aging deterioates the laser performance. Furthermore, the wave length of the optical output is also temperature dependent. To counteract these problems, it has been non-practice to mount the LD on a Peltier-effect thermoelectric cooler with a feedback circuit to stabilize the temperature, and another circuit is included to maintain a constant drive current.
Recentn advances in laser technology using indium phosphide( in P) have resutled in LDs that do not require cooling. This is significant, because coolers are not only expensive but also require considerable power. Cooler-free lasers are very attractive for undersea link applications. There are three major types of LD:
|
|
Fabry-Perot |
|
|
Distributed feedback |
|
|
Distributed Bragg reflector |
Summary of the LD and LED sources. The LED has the following advantages when compared to the LD:
|
|
Higher reliability |
|
|
Simpler drive circuit |
|
|
Lower temperature sensitivity |
|
|
Immunity to reflected light |
|
|
Low cost |
these characteristics make LEDs suitable for short-distance applications. They are particularly attractive for LANs and subscriber loops where economy is a very important factor.
Characteristics of the laser diode.
Optical fiber is gradually finding application in all aspects of telecommunication systems. The three broad categories, which relate to short, medium, and long distance, are referred to as:
1. Local loop (for subscribers) and LANs
The distance, capacity(bitrate), and topology are the primary factors that influence these systems designs and the associated economic viability of constructing and operating them. Also, the extent to which optical fiber is deployed depends on the cost relative to traditional copper-based cable. Optical fiber has already displaced copper for long-haul and inter office traffic, and it is only a matter of time before the local loop becomes fiber. This chapter addresses the various technical factors that need to be taken into account to realize each of the above three system categories.
First, the optical fiber transmission systems technology has been evolving along two different paths:1. Intensity modulated systems
There have been several generations of systems so far. The early systems, before about 1990, did not have the luxury of optical amplifiers, and unregenerated distance was a major cost concern because regenerators were expensive. Optical amplifiers radically changed the landscape of fiber systems, and have now shifted the focus of attention to capacity(that is, bitrate).
First generation systems were primarily limited in regenerators pacing and bitrate by high fiber loss and excessive chromatic dispersion in the fiber caused by the use of LEDs. The development of 1.3um wave length fiber systems produced the improved regenerator spacing and bitrates of the second-generation systems. These systems were still operating with multi-mode fibers, which limited the performance by interference between the propagating modes of the fiber (i.e.,modal dispersion). The move to single-mode fibers operating at 1.3um gave the third generation a very impressive regenerators pacing and bitrate performance. The fourth generation benefited from shifting the operating wave length to 1.55um, which offers the minimum achievable attenuation for silica fibers. TopThe fifth-generation (which appeared in the early 1990s) included optical amplifiers, and today these systems vastly increase the distance between regenerators. Notice the optical amplifier is an optical repeater, and not are generator that involves conversion to the electronic doma in for pulse reshaping and retiming. The sixth-generation is WDM, and these systems became practicable in the mid-1990s. Placing multiple OC-192 bits treamsat different wave lengths on to each fiber allows terabit-per-second through put.
So far, all of the systems described use intensity modulation of the optical transmitter. The seventh generation will use phase-modulation of the optical transmitter, and a coherent detection scheme, which improves the receiver sensitivity. The very narrow spectrum distributed-feedback diodes used in these systems will allow very high operating bitrates. In addition, the coherent optical communication systems will allow FDM, which will give an astounding increase in the channel capacity because of the more efficient use of the available optical band width. The case in favor of using the coherent optical system in future designs appears to be clear-cut, but cheaper WDM designs will provide enough capacity for the near term. Soliton transmission will be the eighth generation, and will no doubt provide enough capacity for very broadband ISDN systems.
Perhaps one important surprise over the past five years has been the pace at which electronics has progressed to enable higher bitrate generation and processing. Only a few years ago, a bitstream of 565Mb/s was considered to be state of the art, but 10-Gb/s bitstreams have appeared remarkably quickly. This design advance has taken considerable pressure off the more exotic and expensive technical solutions for high-capacity systems. In other words, the development and deployment of FDM and soliton-based systems is not a surgent as it was once thought.
In the design of practical systems, there are many variables to take into consideration. First, what is the nature of the interconnection? There are three categories of interconnection: Top
1.Point-to-point(link)
2.Point-to-multi-point(broadcast)
3.Multi-point-to-multi-point(network)
So far, the focus of attention has been on the point-to-point link, and it has been assumed that the ever- increasing hunger for more band-width inevitably leads to the need for coherent detection technology. Although coherent systems are technically more elegant than the structurally simpler direct-detection systems, cost does not yet favor the coherent technology. In fact, the case for employing direct-detection systems is still very strong. When monolithic photonic integrated circuits reach maturity, the cost of coherent detection systems should be low enough to make it the dominant technology for all of the above applications. Until that day, the presently cheaper direct-detection systems will no doubt continue to have wide spread use.
Eight generations of optical fiber
communication systems.
Most optical fiber communications systems presently use the intensity modulation technique. As already stated, this is simply a non- off transmission, whereby the light from the optical source produces 1s or 0s by the light being on or off, respectively. This can be done by directly switching the source on and off, or externally blocking the source to form 1s and 0s. If a laser diode has its current changed from zero to its operating value, the frequency of the output from the laser changes by a small amount. This chirping becomes a very significant problem for coherent detection systems using PM or FM.