FIBER OPTICS. Prof. R.K. Shevgaonkar. Department of Electrical Engineering. Indian Institute of Technology, Bombay. Lecture: 26

Transcription

1 FIBER OPTICS Prof. R.K. Shevgaonkar Department of Electrical Engineering Indian Institute of Technology, Bombay Lecture: 26 Wavelength Division Multiplexed (WDM) Systems Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 1

2 One of the major developments that need to be emphatically mentioned among the modern advancements in fiber optic communication technology is the Wavelength Division Multiplexed (WDM) systems. As the name suggests, multiple wavelengths use the same channel for transfer of data from source to destination. ORIGIN OF WDM SYSTEMS The origin of WDM systems is based on the realisation of underutilization of the full channel capacity of an optical fiber. The figure 26.1 below, shows the two windows of optical communication (already discussed in Transcript 2). Figure 26.1: 2 nd and 3 rd generation windows of Optical communications. The 2 nd and the 3 rd windows of optical communication, put together, provide a total communication bandwidth of about 30THz. With proper technologies, optical fibers have been manufactured which provide such large bandwidths for optical communication. But the optical sources and receivers that are used in an optical system have bandwidth limitations which restrict their maximum modulating frequency to a few GHz only, which is much smaller compared to the enormously large bandwidth provided by the optical fiber. This means that, when an optical fiber is used in a system like the single point-to-point optical communication link, only due to the bandwidth limitation of the transmitter and receiver circuitry, the full channel capacity (bandwidth) of the optical fiber is not utilized and the optical fiber capability remains underutilized. Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 2

3 This realisation led the scientists to use the property of multiplexing, in the optical domain too. The concept was to transmit dissimilar wavelengths carrying information through the same optical fiber. The new optical communication systems possessing this capability, that emerged thereafter, came to be known as wavelength division multiplexed (WDM) systems. Some facts related to WDM systems are bulleted below: WDM systems provide a communication bandwidth ranging from 1.2µm to 1.6µm wherein the optical fiber has a low-loss profile. This bandwidth, in terms of frequency, is equivalent to about 30THz. In terms of channel capacity, the above bandwidth is capable of providing about 300,000 channels, each of speed 10 Mb/s. The reason for using the speed of 10Mb/s is that, even a good quality video transmission would require a speed of about 10Mb/s. That means, if one were to use the entire bandwidth of 30THz of the optical fiber effectively, one could transmit 300,000 video channels simultaneously on the same optical fiber. The above available bandwidth of 30THz may be efficiently utilized in two waystime domain and wavelength domain. However, employment of time domain multiplexing to utilize the entire bandwidth of the optical fiber would rather be impracticable due to the fact that even the most modern pulse generation technology would not be able to generate pulses of width of about a femto-second (10-15 s) which are required to realize a time division multiplexed optical system. The promising option, now left, is the wavelength domain multiplexing which is better known as wavelength division multiplexing, and the systems that possess this feature are called wavelength division multiplexed (WDM) systems. In WDM, the total low-loss bandwidth of the optical fiber ( µm) is sub-divided into multiple channels and each channel having respective carrier signal. Each carrier signal is then modulated at the maximum modulating frequency by the data signal and then these modulated carriers are simultaneously transmitted on the optical fiber. In initial WDM systems, the transmitted channels belonged to two different windows i.e. if one wavelength of transmission was from the 2 nd window, the other was from the 3 rd window of optical communication and vice versa. This was done in order to prevent any possibility of interference among the transmitted channels. However, with time the LASER technology improved tremendously and very precise high performance LASERs with very narrow spectral widths were manufactured. This advancement enabled the simultaneous transmission of two channels belonging to the same window with negligible possibility of interference. This capability of the WDM system caused the optical carriers to be closely spaced in wavelength domain. Due to this dense packing of carrier signals, the new WDM systems came to be known as Dense Wavelength Division Multiplexed (DWDM) systems. The International Telecommunication Union (ITU) has specified standard guidelines for placement of the carrier signal wavelengths within a single communication window. The International Telecommunication Union (ITU) standard (ITU G.692) specifies the inter carrier spacing between two subsequent carrier signals in terms of frequency. Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 3

4 This standard takes into account, all the possibilities of interference between adjacent channels and states that the frequency of the carrier signals must lie on a fixed grid called the ITU grid. ITU G.692 is specified below: ITU G.692 Standard: Reference Frequency: THz ( nm) Fixed Inter-carrier Spacing: 100GHz ( 0.8nm at 1552 nm) One may easily notice that the ITU grid is equi-spaced in the frequency domain but not in the wavelength domain. Also, with two adjacent carrier frequencies 100GHz apart, a 10-20GHz bandwidth signal would not cause any mentionable interference between two channels. With the different carrier wavelengths satisfying the ITU grid, a typical WDM network architecture looks like the one shown below: Figure 26.2: Typical WDM Network Architecture The transmitter (TX) blocks in the figure are the individual optical transmitters. This does not rule out the fact that the TX blocks, in themselves, may be multiplexed output of several other transmitter networks thereby forming some sort of a hierarchical multiplexed architectural system. Also, the transmitter block may consist of time division multiplexed type of system where the data signals to be transmitted is multiplexed in the time domain. The output signals from these transmitters at their corresponding wavelengths ( etc) are then multiplexed in the wavelength domain in accordance to the ITU G.692 standard by the Wavelength Multiplexer (shown in the above figure). The wavelength multiplexer combines all the output signals and combines them to be transmitted along the optical fiber to reach the receiving end. One should, however, note that although the above architecture consists of multiple transmitters and receivers, as far as the type of communication is concerned, it is still a point-to-point link. As per the different design criteria discussed earlier, periodic amplifiers and repeaters are installed along the optical link to establish a secure and satisfactory optical communication link with the minimum possible BER. At the receiver side, the multiplexed transmitted signal is received and then Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 4

5 de-multiplexed by a wavelength de-multiplexer and the respective receivers receive their intended data signals and further processing takes place on these receivers before the signal is actually delivered to the end-user. The above, in principle, is the basic working process of a typical WDM system. However, pure WDM systems have become obsolete with the introduction of the DWDM systems due to a variety of advantages that render DWDM systems far superior over WDM systems. Some of the advantages of DWDM systems are discussed below: 1) Capacity Upgradeable: The significance of this advantage is lies in the fact that new channels can be multiplexed to the already existing (operating) system with only a few minor changes. This facility enables the setting up of new transmitterreceiver pairs as per requirement even after the optical fiber cables have been laid. 2) Transparency: In case of a time division multiplexed (TDM) system, a desired upgrade in the channel capacity can be brought about only after tedious and complicated modification in the manner in which the data flows through the channel. These modifications have to be also made in all of the repeaters installed along the link. For such modifications, precise and accurate timing information about the data has to be available in order to reduce errors. However in DWDM systems, due to the transmission of data via different optical carrier wavelengths, the actual timing information of the data is irrelevant to any up gradation in the channel capacity. Even in the optical repeater/amplifier stations, irrespective of the actual data rate carried by the individual wavelengths, the station just amplifies the signal and re-transmits it. This is the transparency characteristic of the DWDM systems. 3) Wavelength Routing: In packet-data transmission networks, the header bits of each data packet (which determines the address of the destination to the data packet) is read at each router of the network along the path of the data-packet in order to deliver the packet to the appropriate destination through the shortest path. This process uses up valuable transmission time and also there arises possibilities of information being delivered to the wrong destination. However, in DWDM systems, the wavelength itself can be used as the address of a destination and there is no need to read the actual transmission data to route it to the destination. Every destination may be assigned a particular wavelength and the data carried on the wavelength may be appropriately routed to the preassigned destination based on the carrier wavelength. This saves transmission time and also enhances performance of the system. 4) Wavelength Switching: Switching of information signal from one carrier wavelength to another carrier wavelength is also possible in DWDM systems. The need for a switching operation to be initiated occurs whenever there is a non-availability of the required wavelength due to congestion or other network factors. The above characteristics, or rather advantages, have rendered the WDM/DWDM systems far superior as compared to the normal point-to-point communication link. A Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 5

6 variety of devices and systems have been developed which are capable of performing all the above mentioned functions such as wavelength routing, wavelength switching, etc. The next query that comes to the mind is the different components that are required for setting up a WDM/DWDM system for various practical applications. The list below highlights the fundamental components that are required to set up WDM/DWDM system. Dispersion Shifted/Flattened Optical Fiber. Tuneable/Multi-wavelength LASER Optical Source. Broadband Optical Amplifiers. Wavelength Dependent Optical Devices The different background technologies that help in realizing the above components and their intended functions are: Semiconductor Optical Amplifiers (SOA) or Fiber Amplifier (such as Erbium Doped Fiber Amplifier (EDFA), Raman Amplifier) Integrated optical switches / optical couplers Fiber Bragg Gratings (FBG) Arrayed Waveguide Gratings (AWG) Having had the introductory knowledge of the different components required in a WDM system, the subsequent sections shall deal with these components in more detail in order to enhance the understanding of the reader. The first component to be discussed is the optical amplifier- which may either be a Semiconductor Optical amplifier or Erbium Doped Fiber Amplifier (EDFA) or a Raman Amplifier. Let us first acknowledge the actual need for an optical amplifier. For this, let us revert back to the discussions on the Optical Link Design where two different budget calculations were made in order to calculate the inter-repeater distance. For facilitating a comparative discussion let us assume a scenario of a link design as follows: Figure 26.3: Need of Optical Fiber The deterioration of the SNR in the transmission through the optical link is encompassed in the Power budget calculations and the distortion losses in the pulse as it travels through the optical link is taken care of by the rise-time budget calculations. In the figure 26.3 above, a scenario of an optical fiber link has been depicted by the parameters Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 6

7 that have been mentioned under the two heads. Based on these given parameter values, the inter-repeater distances under the two heads have been calculated. From the values of the inter-repeater distance, it is very clear that the lowering of the SNR or the power budget calculations is a prime factor for determining the inter-repeater distance. A repeater is a device which receives the transmitted signal, regenerates it, amplifies it (in the electrical/optical domain) and then re-transmits the signal onto the channel. The repeater is thus a back-to-back integration of a receiver, an amplifier and a transmitter. That is why repeater units are costly too. So, to install repeater units at every ~200Km would not be a cost-effective design. Also, the function of the repeater takes a finite amount of time for the received signal to be amplified and then re-transmitted. However, one must note the fact that the improvement in the SNR of the signal can be brought about by a simple amplification of the transmitted signal which can be done just by installing an amplifier (optical/electrical) unit rather than installing the entire repeater. This fact generates the need for an amplifier unit to be used in the optical fiber link design. In earlier systems, however, repeater units were, in fact, used just for the sake of SNR improvement due to unavailability of appropriate amplifier units. But, the advancements in the optical communications have provided amplifiers that can amplify the signal even in the optical domain and thereby provide very fast and reliable link. Optical Amplifier Technology provides us with two main amplifying strategies which are- semiconductor based amplifier (SOA) of a fiber based optical amplifier (EDFA) as mentioned above. Let us have a quick comparative study between the two in terms of their features and characteristics. Semiconductor Optical Amplifiers (SOA) These amplifiers are compatible both the 1300nm and the 1550nm optical communication windows. SOAs have a high output optical power due to the fact that they are nothing but LASERs which, intrinsically, are optical amplifiers. SOAs have a high coupling loss. Because they are semiconductor based, SOAs can be integrated with other semiconductor based circuitry. They have high non-linearities. Erbium Doped Fiber Amplifier (EDFA) These amplifiers are compatible only in the 1550nm optical communication window. EDFAs can handle high power as well as provide a high gain (25dB). These amplifiers have low coupling losses as well as low noise figures. EDFAs have low cross-talk figures. EDFAs are polarization-insensitive. As can be seen from the above table, EDFAs have number of desirable advantages and so, they have gained high popularity and have become a centre of research over the last decade. The only limitation is the fact that these amplifiers are compatible only in the 1550nm optical communication window. In a generic sense, let us have a basic understanding of an optical amplifier unit which is depicted in the figure 26.4 below: Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 7

8 Figure 26.4: Generic Optical Amplifier As the figure shows, an optical amplifier generally consists of an active medium which is excited by a pump-source. The incident optical input signal is then coupled to this active medium via fiber-to-amplifier couplers and the optical signal gets amplified inside this medium by virtue of stimulated emissions (already discussed earlier) and the amplified optical output s then recoupled to the optical fiber via another fiber-to-amplifier coupler. When the input optical signal travels through the active medium, the power from the pump gets transferred to the signal power and thus we obtain an amplified optical output. The energy level structure of the active medium, in fact, has to be such that forbidden-energy gap corresponds to the energy of the incoming photons, so that the stimulated emission process is initiated. Hence, the first step in the construction of an optical amplifier is to identify suitable materials that can be used as active media in the amplifier either directly or after suitable treatment to a basic material. The latter, in fact, is done in the construction of an EDFA in which the core of a normal optical fiber is appropriately doped with erbium (Er 3+ ) atoms and the energy level diagram of the doped region of the optical fiber core thus looks like: Figure 26.5: Erbium Energy Level Diagram Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 8

9 As can be seen from the above diagram, the energy difference between the metastable and the ground states corresponds to about 1550 nm and so the EDFA are said to operate in the 1550nm optical communication window. The metastable state acts as the lasing level in the EDFA where the electrons await stimulation from the incident optical photons to undergo stimulated emission so as to provide an amplified coherent optical output. However, due to the wide nature of the metastable band (1480nm-1560nm), two different types of pumping wavelengths may be possible viz. 980nm pump and 1400nm pump as shown in the above figure. But the use of a 1400nm pump poses difficulties such as higher noise and lesser separation between pump and signal wavelengths thereby creating the possibility of interference. Hence a 980nm pump which has lower noise is generally preferred. The pumped electrons to the pump band then undergo a rapid non-radiative decay to the metastable band. The time constant of this decay is of the order of micro-seconds. Once in the metastable band, the electrons in the higher metastable levels further decay down to the lower metastable levels where there are more number of energy levels available and here they wait to enrol in the process of stimulated emission. Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 9