**1. Introduction**

Any form of telecommunication system that having light as the transmission medium is known as optical communication. Optical communication system consists of a transmitter, channel and a receiver. The transmitter will **encode** a message into an optical signal, the chan‐ nel will carry the signal to its destination, and the receiver will reproduce the message from the received optical signal.

© 2017 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Radio over fiber (RoF) is a technology where RF signal modulates light and then transmitting it over an optical fiber link. Both wireless network and optical are supported by this technol‐ ogy. It is essential for communication system to have high capacity and subcarrier frequency since wireless signal sometimes intend to loss channel at the time of data transmission [1].

RoF is very convenient system since it is less costing and power consumption. This is because RoF lets the electrical signal modulates the optical source and after that the optical signal will travel along the optical fiber to the remote station. When the RF signal is modulated straight to the optical link, the power consumption drops while the antenna side has high frequency radio carriers. The reduction of cost in RoF technology can be explained in two ways. The first one is central station (CS), which provides resources that can be shared by variety of base stations (BS), and secondly, BS only executes simple function. Furthermore, the BS is in a small size and less cost consuming.

Basically, in this technology, central station (CS) is connected with many base stations (BS) by using optical fiber. BS only functions as a converter of optical signal into a wireless signal and vice versa, while at CS, all process involving modulation, demodulation, coding and routing are executed [2]. By using high linear optic link, RoF system distributes the RF signal between CS and BSs. **Figure 1** shows the basic construction of radio over fiber network transmission.

At the transmitter side, data or information from internet or other CS are fed onto modem in the CS during downlink process. Optical signal from the optical source is modulated by the RF signal. After that, the modulated signal will pass through optical fiber toward the BSs. As soon as the signal reaches the BS, photodetector (PD) will function as a detector to detect the modulated optical signal. The PD will also recover the signal before the signal is transmitted through antenna of the BS toward the mobile host. With the similar concept, reverse process is executed for uplink process between mobile host and CS. When the BS receives the signals, optical signal from the LD will be modulated to amplified and transmitted straight toward CS.

In order to determine either the receiver of the RoF in a good quality or not, we can measure and analyze the bit error rate (BER), Q factor value and the eye opening of the resulting result. Bit error rate is the number of received bits of a data stream over a communication channel that has been altered due to noise, interference and distortion orbit synchronization errors. International Telecommunication Union (ITU) has stated that the minimum value of BER of RoF must be below than 10<sup>−</sup><sup>9</sup> . Basically, the value of BER is depending on the measurement time and factors lead to the error. The value of Q depends on the value of BER. There are sev‐ eral ways to determine either our obtained q factor is suitable for our BER value or not. The easiest way is to determine it from BER and Q factor graph. **Figure 2** shows the graph that indicates the relationship between the value of BER and Q factor.

**Figure 1.** Basic construction of radio over fiber.

**Figure 2.** BER versus Q factor.

Radio over fiber (RoF) is a technology where RF signal modulates light and then transmitting it over an optical fiber link. Both wireless network and optical are supported by this technol‐ ogy. It is essential for communication system to have high capacity and subcarrier frequency since wireless signal sometimes intend to loss channel at the time of data transmission [1].

RoF is very convenient system since it is less costing and power consumption. This is because RoF lets the electrical signal modulates the optical source and after that the optical signal will travel along the optical fiber to the remote station. When the RF signal is modulated straight to the optical link, the power consumption drops while the antenna side has high frequency radio carriers. The reduction of cost in RoF technology can be explained in two ways. The first one is central station (CS), which provides resources that can be shared by variety of base stations (BS), and secondly, BS only executes simple function. Furthermore, the BS is in a small size and less

Basically, in this technology, central station (CS) is connected with many base stations (BS) by using optical fiber. BS only functions as a converter of optical signal into a wireless signal and vice versa, while at CS, all process involving modulation, demodulation, coding and routing are executed [2]. By using high linear optic link, RoF system distributes the RF signal between CS and BSs. **Figure 1** shows the basic construction of radio over fiber network transmission. At the transmitter side, data or information from internet or other CS are fed onto modem in the CS during downlink process. Optical signal from the optical source is modulated by the RF signal. After that, the modulated signal will pass through optical fiber toward the BSs. As soon as the signal reaches the BS, photodetector (PD) will function as a detector to detect the modulated optical signal. The PD will also recover the signal before the signal is transmitted through antenna of the BS toward the mobile host. With the similar concept, reverse process is executed for uplink process between mobile host and CS. When the BS receives the signals, optical signal from the LD will be modulated to amplified and transmitted straight toward CS. In order to determine either the receiver of the RoF in a good quality or not, we can measure and analyze the bit error rate (BER), Q factor value and the eye opening of the resulting result. Bit error rate is the number of received bits of a data stream over a communication channel that has been altered due to noise, interference and distortion orbit synchronization errors. International Telecommunication Union (ITU) has stated that the minimum value of BER of

time and factors lead to the error. The value of Q depends on the value of BER. There are sev‐ eral ways to determine either our obtained q factor is suitable for our BER value or not. The easiest way is to determine it from BER and Q factor graph. **Figure 2** shows the graph that

indicates the relationship between the value of BER and Q factor.

. Basically, the value of BER is depending on the measurement

cost consuming.

160 Optical Fiber and Wireless Communications

RoF must be below than 10<sup>−</sup><sup>9</sup>

**Figure 1.** Basic construction of radio over fiber.

From the graph, we can see that the value of Q factor increases when the BER decreases. From the graph also, we could see that the value of Q factor for 10<sup>−</sup><sup>9</sup> of BER is approximately 6. Thus, RoF system should have a value of Q factor more than 6 in order to obtain a very good performance of its receiver.

The receiver performance of RoF can also be measured by analyzing the eye diagram of the result after the design has been simulated. Eye diagram analyzer shows multiple traces of a modulated signal to produce eye diagram. This eye diagram is an oscilloscope display where the eye pattern diagram corresponds to minimal signal distortion due to intersymbol interference (ISI) and noise appear in the system [3]. The measurements of eye pattern are made based on time domain. The effect of waveform distortion will appear on the screen of regular BER test equipment. There are many information that could be obtained from eye pattern display. Time interval over which signal that has been reach at the receiver could be sampled without any error causes by ISI could be defined by the width of the eye opening. Noise margin can be determined by looking at the height of eye opening at specified sam‐ pling time.

## **2. Overview of radio over fiber**

In 1990, RoF is firstly demonstrated for mobile telephone services or cordless. In this technology, highly linear optic fiber links are used to connect CS and BSs so that RF signal distribution between CS and BSs can be done. Processes involving modulation, demodulation, coding and routing are executed mostly at CS. That means, the only processes occur in BSs only converting optical signal to electrical signal or vice versa. A lot of studies and research have been done just to investigate the limitation and generate new idea to increase the performance of RoF technologies.

Basic RoF consists of all hardware that enable to foist RF signal on an optical carrier at the trans‐ mitter side. It also needs fiber optic link to distribute the signal from CS to BSs. At the receiver side, RoF required all the hardware needed to recover the signal from the optical carrier.

#### **2.1. Optical transmission link**

#### *2.1.1. Optical fiber*

Optical fiber is a platform or a medium to carry information in the form of light from a point to another point. Functioning as a waveguide that will allow the propagation of light, a fiber is a thin filament of glass. One of the advantages of fiber is it provides a path for light with some losses due to the concept of total internal reflection.

In communication system, there are three types of optical fiber, which are step‐index mul‐ timode, step‐index single mode and graded index. Step‐index multimode fiber is measured from cladding to the core to cladding as it serves an index of refraction profile that steps from low to high to low. For step index single mode, only one path is allowed for the light to travel in the fiber. Graded index provides large core diameter and higher bandwidth of single‐mode fiber.

Optical fiber provides two regions that have low attenuation. First region is at approximately 1300 nm, which has attenuation less than 0.5 dB/km and bandwidth 25 THz. Second region is at approximately 1550 nm, which has attenuation less than 0.2 dB/km and also has band‐ width 25 THz [4]. This combination of two regions will make the total bandwidth as much as 50 THz. Due to the both low attenuation regions, signal loss in data transmission will be very small. Thus, we do not need a lot of amplifiers and repeaters.

#### *2.1.2. Attenuation, dispersion and nonlinearities in fiber*

In a fiber optic, attenuation can affect the signal power during the propagation of the signal through distances. Attenuation is a must aspect need to be considered in order to determine the longest distance that signal can go for a given sensitivity of the receiver and the power of the transmitter.

Widening of pulse duration when it propagates through a fiber is known as dispersion. When one pulse starts widening, the pulse is interfering with another pulse besides it. Thus, inter‐ symbol interference (ISI) can happen. Because of this phenomenon, the pulse spacing and the maximum transmission will be limited. There are several types of dispersions and one of them is intermodal dispersion. Intermodal dispersion can happen when several modes of similar signal travels at different velocities through the fiber. Single‐mode fiber would not have this kind of dispersion.

Chromatic dispersion is another type of dispersion. Most system will have this type of disper‐ sion since there is no laser able to create a signal that having a single wavelength. Wavelength will be functioning as an index of refraction in dispersive medium. Certain wavelength will travel more fast than other wavelength if the signal that being transmitted has more than one wavelength [5]. Waveguide dispersion might be happen when the propagation of not similar wavelengths depends on characteristics of the waveguide such as indices and shape of fiber core and cladding. Chromatic dispersion is almost 0 in single‐mode fiber at 1330 nm. About 1330 nm is also a low attenuation region, and thus, fibers with 0 dispersion can be achieved by using advance techniques such as dispersion shifting.

Nonlinearities may cause an attenuation, distortion and cross‐channel interference. Its effect is able to affect the performance of wavelength division multiplexing (WDM) system. In WDM system, nonlinear effects can affect the spacing between similar wavelength channels, limit the maximum power of any channel and able to limit the maximum bit rate.

#### **2.2. Optical transmitter**

#### *2.2.1. Optical sources*

coding and routing are executed mostly at CS. That means, the only processes occur in BSs only converting optical signal to electrical signal or vice versa. A lot of studies and research have been done just to investigate the limitation and generate new idea to increase the performance

Basic RoF consists of all hardware that enable to foist RF signal on an optical carrier at the trans‐ mitter side. It also needs fiber optic link to distribute the signal from CS to BSs. At the receiver side, RoF required all the hardware needed to recover the signal from the optical carrier.

Optical fiber is a platform or a medium to carry information in the form of light from a point to another point. Functioning as a waveguide that will allow the propagation of light, a fiber is a thin filament of glass. One of the advantages of fiber is it provides a path for light with

In communication system, there are three types of optical fiber, which are step‐index mul‐ timode, step‐index single mode and graded index. Step‐index multimode fiber is measured from cladding to the core to cladding as it serves an index of refraction profile that steps from low to high to low. For step index single mode, only one path is allowed for the light to travel in the fiber. Graded index provides large core diameter and higher bandwidth of single‐mode

Optical fiber provides two regions that have low attenuation. First region is at approximately 1300 nm, which has attenuation less than 0.5 dB/km and bandwidth 25 THz. Second region is at approximately 1550 nm, which has attenuation less than 0.2 dB/km and also has band‐ width 25 THz [4]. This combination of two regions will make the total bandwidth as much as 50 THz. Due to the both low attenuation regions, signal loss in data transmission will be very

In a fiber optic, attenuation can affect the signal power during the propagation of the signal through distances. Attenuation is a must aspect need to be considered in order to determine the longest distance that signal can go for a given sensitivity of the receiver and the power of

Widening of pulse duration when it propagates through a fiber is known as dispersion. When one pulse starts widening, the pulse is interfering with another pulse besides it. Thus, inter‐ symbol interference (ISI) can happen. Because of this phenomenon, the pulse spacing and the maximum transmission will be limited. There are several types of dispersions and one of them is intermodal dispersion. Intermodal dispersion can happen when several modes of similar signal travels at different velocities through the fiber. Single‐mode fiber would not

of RoF technologies.

*2.1.1. Optical fiber*

fiber.

the transmitter.

have this kind of dispersion.

**2.1. Optical transmission link**

162 Optical Fiber and Wireless Communications

some losses due to the concept of total internal reflection.

small. Thus, we do not need a lot of amplifiers and repeaters.

*2.1.2. Attenuation, dispersion and nonlinearities in fiber*

The most common of light sources used in fiber optic communications are laser diode and light‐emitting diode (LED). The benefit of these devices is both have output power for wide range applications. The power also can be directly modulated where the input current is var‐ ied to the devices. The efficiency is also high, and they are compatible with the optical fiber.

The diffrence between LED and laser diode is that the laser diodes gives a coherent output where the optical energy is produced in an optical resonant cavity. **Figure 3** shows a basic structure of a laser.

Two mirrors in the laser will form a space between both of them called cavity, a lasing medium that occupied the cavity and a device for excitation. Lasing medium will receive current by the excitation devices and will produce a photon of light. The photon will reflect off the mir‐ rors at both end of the cavity and will go through the medium again.

**Figure 3.** Laser structure.

#### *2.2.2. Optical modulation and line coding*

To transmit information across an optical fiber, the data are compulsory to be encoded or modulated onto a laser signal to allow the data being transported through the optical. There is variation of method for modulation. This is including an analog techniques such as frequency modulation (FM), phase modulation (PM) and amplitude modulation (AM). For digital sig‐ nal, method of modulation includes amplitude shift keying (ASK), phase shift keying (PSK) and frequency shift keying. Among these techniques, binary ASK is the most preferable due to its simplicity. In the systems implementing ASK, the laser is switched on and of to achieved modulation techniques [6]. Current optical communication system also reported on the usage of on‐off‐Keying (OOK) and DPSK modulations.

To transport digitized information in a communication link, format of transmission the sig‐ nal must be considered. The signal format is so important since the receiver needs to extract accurately the timing information from the incoming signal. Line coding has several principal functions and one of it is to minimize the errors which causes by noise or any interference effects in the bit stream. The easiest method for encoding data or information is unipolar return to zero (NRZ) code. Logic 1 is represented by a light source that fills the whole bit period, while logic 0 is represented by no pulse transmitted. The process turns the voltage on and off and that's why it is known as (ASK) or on‐off keying (OOK). It is essential for NRZ to have a minimum bandwidth and NRZ must be simple to generate and decode.

## **2.3. Optical receiver**

Optical receiver consists of signal processing circuitry, an amplifier and a photodetector [7]. The first thing receiver did when received a signal was converting the optical signal into an electrical signal. After that, the signal will be amplified to an optimum level so that the fol‐ lowing process can be done. It is essential to determine and predict the performance of the system based on mathematical models of many receiver stages in order to design a receiver. Also, when designing a receiver, noises and distortions contributed by component in every stage must be considered. Plus, the receiver must have the ability in detecting weak or dis‐ torted signals, ability in making a decision on type of received signal and ability to reshaping the distortion signal. This is why it is more complicated compared to process of designing a transmitter. Bit error rate is the most important criteria in measuring RoF system. Other than that are Q factor and the opening of eye diagram.

The height of the eye diagram opening indicated the level of signal distortion. The upper level of the eye represents binary '1' and the bottom level represents binary '0'. Higher eye opening height is desirable, as this indicating that the binary '1' and '0' can be distinguished well. The height of eye opening at a specified time corresponds to the noise margin achieved.

#### *2.3.1. Photodetector*

It is compulsory for a receiver to have a device that can interpret the information in the opti‐ cal signal. Photodetector is a device that can convert the incoming photonic stream into a stream of electrons. When the optical signals reached the receiver, it is in weak and distort condition after undergoes the optical fiber. Thus, photodetector should be sensitive to the emission wavelength range of the optical sources being used. Photodetector also must have an addition of noise to the system, and most importantly, it has to response fast to handle the target data rate. Also, photodetector must be immune to the change of temperature and must be compatible with the physical dimensions of optical fiber. After the photodetector changes the optical signal into an electrical signal, the signal will be amplified and will undergo the threshold device. To determine either the bit is 0 or 1, the presence of light is referred to during the bit duration. It depends on either the electron stream is below or above a certain threshold.

PIN photodiode is the most usable semiconductor photodetector. The structure of this device as shown in **Figure 4** consists of p and n regions. Both regions are separated by intrinsic (i) region [8]. For normal operation, an optimum reverse‐bias voltage is supplied across the pho‐ todiode to allow the intrinsic region completely depleted of carries. This will cause the n and p carrier concentrations that become less than impurity concentrations in the region. When a photon flux Φ penetrated into the device, the flux is absorbed.

In PIN photodetector, the light absorption causes the formation of electron‐hole pairs. Then, the hole and electron are drifted to the opposite direction, causing the flow of current. More current flows as more light enter the photodetector, which giving rise to the number of electron‐hole pairs.

When the energy of an incident photon is higher than or equal to the semiconductor's band‐ gap energy, the photon excite an electron from valence band to the conduction band by give up its energy. The absorption process will form an electron‐hole pairs known as photo car‐ riers. Normally, the photodetector is designed so that those carriers are mainly generated in the depleted intrinsic region. This region has the most absorption of incident light. The large amount of electric field will make the carriers separating between each other and will be collected across the reverse biased junction. Thus, current flow will increase in the external circuit where every carrier pair generated has one electron flowing.

**Figure 4.** PIN photodiode structure.

*2.2.2. Optical modulation and line coding*

164 Optical Fiber and Wireless Communications

of on‐off‐Keying (OOK) and DPSK modulations.

that are Q factor and the opening of eye diagram.

**2.3. Optical receiver**

*2.3.1. Photodetector*

To transmit information across an optical fiber, the data are compulsory to be encoded or modulated onto a laser signal to allow the data being transported through the optical. There is variation of method for modulation. This is including an analog techniques such as frequency modulation (FM), phase modulation (PM) and amplitude modulation (AM). For digital sig‐ nal, method of modulation includes amplitude shift keying (ASK), phase shift keying (PSK) and frequency shift keying. Among these techniques, binary ASK is the most preferable due to its simplicity. In the systems implementing ASK, the laser is switched on and of to achieved modulation techniques [6]. Current optical communication system also reported on the usage

To transport digitized information in a communication link, format of transmission the sig‐ nal must be considered. The signal format is so important since the receiver needs to extract accurately the timing information from the incoming signal. Line coding has several principal functions and one of it is to minimize the errors which causes by noise or any interference effects in the bit stream. The easiest method for encoding data or information is unipolar return to zero (NRZ) code. Logic 1 is represented by a light source that fills the whole bit period, while logic 0 is represented by no pulse transmitted. The process turns the voltage on and off and that's why it is known as (ASK) or on‐off keying (OOK). It is essential for NRZ to

Optical receiver consists of signal processing circuitry, an amplifier and a photodetector [7]. The first thing receiver did when received a signal was converting the optical signal into an electrical signal. After that, the signal will be amplified to an optimum level so that the fol‐ lowing process can be done. It is essential to determine and predict the performance of the system based on mathematical models of many receiver stages in order to design a receiver. Also, when designing a receiver, noises and distortions contributed by component in every stage must be considered. Plus, the receiver must have the ability in detecting weak or dis‐ torted signals, ability in making a decision on type of received signal and ability to reshaping the distortion signal. This is why it is more complicated compared to process of designing a transmitter. Bit error rate is the most important criteria in measuring RoF system. Other than

The height of the eye diagram opening indicated the level of signal distortion. The upper level of the eye represents binary '1' and the bottom level represents binary '0'. Higher eye opening height is desirable, as this indicating that the binary '1' and '0' can be distinguished well. The

It is compulsory for a receiver to have a device that can interpret the information in the opti‐ cal signal. Photodetector is a device that can convert the incoming photonic stream into a

height of eye opening at a specified time corresponds to the noise margin achieved.

have a minimum bandwidth and NRZ must be simple to generate and decode.

#### *2.3.2. Optical amplifier*

Optical amplifiers can contribute a lot to long haul or local networks even optical signal still can transmit a long distance without amplifiers. There are several techniques of optical ampli‐ fication. 1R (regeneration) techniques provide a booster to power up the signal. It does not restore the shape and the timing of signal. It also serves a total data transparency. 2R (regen‐ eration & reshaping) amplification is a technique where the optical signal will be transformed into electrical signal before it directly used to modulate a laser.

3R (regeneration, reshaping, relocking) techniques change the data stream into electronic signal and after that retransmit the signal optically to amplifies the signal [9]. Noise can be eliminated much through the reshaping process where the signal produces back the original pulse shape of each bit. Mostly, reshaping works on digital signal but for some cases it may also works on analog signal but relocking does not work on analog modulated sign.

Basically, there are three main types of optical amplifiers. They are semiconductor optical amplifier (SOA), doped fiber amplifier (DFA) and Raman amplifier. The main function of all amplifiers is to boost the power level of incident light through optical power transfer process. Also, this process can be done through a stimulated emission. Using the concept of laser diode, the mechanism of SOA and DFA is creating the population inversion which is needed for simulated emission. Optical amplifier does not have the ability to generate coherent opti‐ cal output since it does not have optical feedback which is compulsory for lasing. Thus, opti‐ cal amplifier can only increase the signal levels.

The amplifier gained an energy from a pump, which is an external source as shown in **Figure 5**. Technically, the external sources are supplying an energy to the electrons in the active medium. This will cause the electrons increased and served a population inversion. The excited electrons will drop to lower level because the electrons are triggered by the incoming photon signal through the process of simulated emission. The output of the signal will be amplified since one incoming photon stimulates a cascaded effect in which way equal energy of photons is emitted by a lot of excited electrons when they hit the ground state.

Advantage of SOA is that this amplifier can be implemented if both signal processing and switching functions were call in optical networks. The drawbacks of SOA is it has rapid gain response that will cause the gain at specific wavelength fluctuate with signal rate for speed up

**Figure 5.** Signal amplification by an optical amplifier.

to some Gb/s. The overall gain also can be affected, and this will cause the signal gain of other wavelengths fluctuate. Thus, cross‐talk effect can occur.

*2.3.2. Optical amplifier*

166 Optical Fiber and Wireless Communications

Optical amplifiers can contribute a lot to long haul or local networks even optical signal still can transmit a long distance without amplifiers. There are several techniques of optical ampli‐ fication. 1R (regeneration) techniques provide a booster to power up the signal. It does not restore the shape and the timing of signal. It also serves a total data transparency. 2R (regen‐ eration & reshaping) amplification is a technique where the optical signal will be transformed

3R (regeneration, reshaping, relocking) techniques change the data stream into electronic signal and after that retransmit the signal optically to amplifies the signal [9]. Noise can be eliminated much through the reshaping process where the signal produces back the original pulse shape of each bit. Mostly, reshaping works on digital signal but for some cases it may

Basically, there are three main types of optical amplifiers. They are semiconductor optical amplifier (SOA), doped fiber amplifier (DFA) and Raman amplifier. The main function of all amplifiers is to boost the power level of incident light through optical power transfer process. Also, this process can be done through a stimulated emission. Using the concept of laser diode, the mechanism of SOA and DFA is creating the population inversion which is needed for simulated emission. Optical amplifier does not have the ability to generate coherent opti‐ cal output since it does not have optical feedback which is compulsory for lasing. Thus, opti‐

The amplifier gained an energy from a pump, which is an external source as shown in **Figure 5**. Technically, the external sources are supplying an energy to the electrons in the active medium. This will cause the electrons increased and served a population inversion. The excited electrons will drop to lower level because the electrons are triggered by the incoming photon signal through the process of simulated emission. The output of the signal will be amplified since one incoming photon stimulates a cascaded effect in which way equal energy

Advantage of SOA is that this amplifier can be implemented if both signal processing and switching functions were call in optical networks. The drawbacks of SOA is it has rapid gain response that will cause the gain at specific wavelength fluctuate with signal rate for speed up

of photons is emitted by a lot of excited electrons when they hit the ground state.

also works on analog signal but relocking does not work on analog modulated sign.

into electrical signal before it directly used to modulate a laser.

cal amplifier can only increase the signal levels.

**Figure 5.** Signal amplification by an optical amplifier.

For DFA, the length of fiber will be doped with an element that able to amplify the light [10]. The most common element is erbium. The main advantage for this type of amplifier is it can pump the devices at some different wavelengths. DFA also has small coupling loss and low dependence of gain on light polarization. Plus, DFA is immune from interference effects since the gain responses are constant for signal modulations greater than few kilohertz.

Raman amplifier has a transfer of optical power form high power pump wavelength to light wave signals at longer wavelengths. Raman amplification does not need process of popula‐ tion inversion. It is based on stimulated Raman scattering (SRS) which is a nonlinear effect. This effect normally occurs in fibers at high optical powers. Raman gain mechanism was achieved through a discrete amplifier. This type of amplifier can be used in any wavelength band since the gain is based on SRS where the induced transfer of optical power from shorter pump wavelengths to longer signal wavelengths.
