DWDM Networks When Using Raman Configurations with DCF Fibers

In this paper, the basic theoretical foundations regarding a DWDM [Dense Wavelength Division Multiplexing] network when using Raman configurations with DCF [Dispersion Compensation Fiber] are studied, through the revision of: linear effects, such as noise, loss, chromatic dispersion and PMD [Polarization Mode Dispersion]; and nonlinear effects, such as Self-Phase Modulation [SPM], Cross-Phase Modulation [XPM] and Four-Wave Mixing [FWM], compared to Optical Monitoring Parameters [OPM].


I. Introduction
At present, data transport networks must be able to offer high levels of bandwidth due to the appearance of numerous applications such as video conferencing, growing internet traffic, real-time operations, high-speed services, etc.The constant growth of this demand has brought existing technologies to their limits; a situation that leads to require new concepts and innovations that allow the implementation of more robust networks, with greater capacity and higher transmission speeds.These needs that have arisen as a result of technological development have made optical fiber an attractive and highly projected medium to support the resources demanded by services of users, in terms of capacity and scope (García, 2011).
To increase the information transmission capacity, optical fiber employs Wavelength Division Multiplexing [WDM] techniques, which reduce network costs; however, such systems have limitations when used over long distances and at higher speeds, and are subject to optical degradations which generate undesired modifications in the signal.Dense Wavelength Division Multiplexing [DWDM] transmission systems represent an alternative to achieve transmissions over longer distances; the system in question is characterized mainly by providing a spacing between smaller wavelengths, increasing its transport capacity (Rec.ITU-T G.694.  and requiring, at the same time, the incorporation of new technologies and mechanisms, in order to minimize the incidence of linear and nonlinear effects on system performance. Considering that Raman distribution amplifiers are a promising technology, by reducing the power of the input signal, without sacrificing the Optical Signal to Noise Ratio [OSNR] -the performance parameter that measures how much it degrades the optical signal that is transported within the system-after the transmission (Liu, 2002), using the Stimulated Raman Scattering [SRS] as the gaining mechanism of the amplifier (Rocco et al., 2009), and the Dispersion Compensation Fiber [DCF], allow to counteract the degradations generated by the chromatic dispersion thanks to the contribution of its properties, such as the negative dispersion index, attenuation and figure of merit, among others (Hoyos & Vélez, 2013).Having said that, this article focuses on performing an analysis in DWDM networks when using Raman configurations with DCF fibers; for this, it is initially necessary to review basic theoretical fundamentals, bibliographic revision and works related to the subject in reference.

lI. General considerations, regarding WDM and DWDM
WDM is a multiplexing technique for optical networks that allows multiplying the capacity of each link by transmitting information simultaneously in a single fiber, using different wavelengths, and in this way it manages to optimize the fiber bandwidth and higher transmission rates.However, one of the weaknesses of this system is that its channels are not closely linked in the frequency domain, which reduces the efficiency in the transmission medium; additionally, there may be difficulties in responding to different optical degradations, such as linear and nonlinear effects, which become a limiting of the link distance and system speed (Alvarado, 2009).
On the other hand, due to the evolution and commercial offer of computer services, such as videoconferences, multimedia applications and telemedicine, among others, the DWDM provides an alternative to increase the bandwidth, as it allows to adjust in a single fiber around a hundred of wavelengths, without mutual interference; each wavelength represents an optical channel, over which a service is transported.The functions of DWDM are multiple, including: amplifying all wavelengths without having to convert to electrical signals, having independence between protocols and capacities, and transporting signals with different speeds; in spite of that, one of the main features of DWDM is that it can send 32/40/64/80/96 multiplexed virtual channels on a single optical fiber, with a tendency to increase thanks to several modern techniques used by suppliers (Meléndez, 2013).At present, up to 320 optical channels can be achieved transporting any transmission format, so it is possible to use different wavelengths to send analog or digital information (Sembroiz, 2013).DWDM systems require higher wavelength accuracy and filter throughput than conventional WDM systems; this is because a small deviation in the center of the wavelength of one of the lasers can distort the signal of the adjacent channel, thus, the stability of the lambdas or channels is of utmost importance (Ferrín, 2014).
Although DWDM has become an innovative, effective and determinant technique for the transport of information in optical systems, certain conditions must be taken into account: first of all, it is a technology that is at a midpoint of its development and is very dependent on the characteristics of the fibers; in second place, it is required high-quality optical components such as temperature-controlled lasers and cooling systems, more developed multiplexers and demultiplexers, which imply an increase in the total cost of the network to be installed (García, 2006).Finally, it is essential to mention that the fundamental difference between WDM and DWDM is the spacing between channels; in DWDM they are closer, therefore it has more capacity and makes use of a larger spectrum margin (García, 2006).

lII. Stimulated Raman scattering or SRS
The origin of the Raman effect occurs when a high-intensity monochromatic light beam has an impact on the vibrations of the molecules and atoms of the material that forms an optical fiber; these vibrations characteristic of the material medium are called optical phonons.The result of the interaction causes that part of the energy is going to be absorbed by the medium and is going to be presented like a phonon that remains within the system; the rest of the energy leaves, generally, as a new photon of lower energy (less frequency and consequently, greater wavelength) than the incident photon; while the resulting photon is called Stokes, it can travel in the same direction of the incident beam (co-propagation) or in the opposite direction (counter-propagation).On the contrary, if the energy is higher (higher frequency) the anti-Stokes dispersion occurs (Sembroiz, 2013;Álvarez, Hernández, & Quiroz, 2007;Hernández, 2011).The representation of the SRS process, in terms of electronic levels, can be seen in Figure 1.
Figure 1 shows that a photon with energy Wp (orange signal) is absorbed by the medium and makes a transition from the ground state to a virtual state of excited energy; after some time, part of that energy is transferred and the rest gives rise to a new energy photon Ws (red signal).If the final state of the transferred energy is located at a higher level than the initial state, the photon emission occurs at a lower frequency (lon-instalar (García, 2006).Finalmente, resulta imprescindible mencionar que la diferencia fundamental entre WDM y DWDM es el espaciamiento entre canales, en DWDM son más cerrados, por lo tanto, tiene más capacidad y hace uso de un mayor margen del espectro (García, 2006).

IV. Fibras compensadoras de dispersión
Para combatir la CD (fenómeno lineal que produce ensanchamiento de los pulsos transmitidos), se han desarrollado las Fibras Compensadoras de Dispersión (DCF, Dispersion Compensating Fiber), las cuales son ampliamente usadas en los modernos enlaces ópticos; su característica principal es presentar una alta dispersión negativa, con valores que oscilan entre -100 y -300 ps/ nm.km (el primero es el de mayor uso comercial), lo que resulta una pendiente negativa para compensar la dispersión positiva de la fibra en las bandas C y L; tales bandas, funcionan empleando pequeños tramos de fibras de dispersión cromática elevada y de signo contrario a la que ha introducido la fibra principal; de esta manera, se espera que de tramo en tramo, la dispersión cromática total sea prácticamente nula, lo que evita la excesiva deformación de los pulsos de luz y la distorsión de la señal (Hoyos & Vélez, 2013;Mena & Mendoza, 2009;Chomycz, 2009).ger wavelength) than the incident photon; in this case, the photons emitted produce the Stokes wave.On the other hand, if the final state of the transferred energy is located at a lower level than the initial state, the photon emission occurs at a higher frequency (shorter wavelength) than the incident photon; in this case, the wave that forms the photons emitted is an anti-Stokes wave (Hernández, 2011;Carrasco, 2007); as a consequence of this process, the initial photon ceases to exist, so Raman amplifiers must be optically bombarded to produce gain.
Raman amplification uses SRS as the amplifier gain mechanism (Rocco et al., 2009).To obtain a Raman amplifier using the optical fiber as a gain medium, the pumping signal and the useful signal must be transmitted on the same fiber, regardless of the direction of propagation; thus, the SRS is produced.The energy is transferred from the pumping to the signal by the Raman process, stimulated while both radiations are propagated on the fiber (Álvarez et al., 2007).Figure 2 shows how a fiber can be used as a Raman amplifier, where pumping and signal are injected into the fiber through a fiber coupler.

IV. Dispersion compensating fibers
To combat CD (linear phenomenon that produces broadening of transmitted pulses), Dispersion Compensating Fibers [DCF] have been developed, which are widely used in modern optical links; the main characteristic is to present a high negative dispersion, with values ranging from -100 to -300 ps/nm.km(the first one is the one with the highest commercial use), which result in a negative slope to compensate the positive fiber dispersion in the C and L bands; such bands operate using small sections of high chromatic dispersion fibers and opposite to the one introduced by the main fiber; in this way, it is expected that the whole chromatic dispersion will be practically null, which avoids the excessive de-   (Álvarez et al., 2007) / Diagrama de un amplificador Raman (Álvarez et al., 2007) Muñoz, G. (2017).
The most relevant properties of the DCF are attenuation, dispersion coefficient, noise figure and effective area (Hoyos & Vélez, 2013): • the attenuation is loss of the power level of the signal, changes linearly with the distance of the link, that is, the greater fiber length, the greater attenuation and vice versa; • the scattering coefficient is a linear phenomenon that deforms transmitted pulses, if it is excessive, it will manifest in the overlap of 1s and 0s, which generates distortion and errors in decoding; this value is characterized by being negative and being one of the most important parameters, since it allows to counteract the total dispersion accumulated in the link of the conventional fiber; • the figure of merit defines the relationship between dispersion and DCF attenuation; and • the effective area is the core section of the fiber through which all light power passes (Gaxiola, 2005).
There are three basic schemes to compensate the CD: • pre-compensation scheme, where the CD is compensated before the transmission fiber section; • post-compensation scheme, where the CD is compensated after the fiber; and • symmetrical compensation scheme, in which DCF sections are used, before and after the transmission fibers.
Figure 3 shows the proposed schemes.
In the development of the present work the post-compensation technique is taken into account in order to reduce the BER (Bit Error Rate) and to witness in a more significantly way the nonlinear effects, due to a lower power level decrease compared to the pre-compensation technique (Hoyos & Vélez, 2013;Criollo & Lasso, 2014).Likewise, the impulse that the signal must have to be amplified by Raman requires that there are no connections that generate considerable losses between the transmitter and the optical fiber that will be used as an amplification medium.
Las propiedades más relevantes de la DCF son: la atenuación, el coeficiente de dispersión, la figura de ruido y el área efectiva (Hoyos & Vélez, 2013): • la atenuación es pérdida del nivel de potencia de la señal, cambia linealmente con la distancia del enlace, es decir, a mayor longitud de la fibra mayor atenuación y viceversa; • el coeficiente de dispersión es un fenómeno lineal que deforma los pulsos transmitidos, si es excesivo, se manifestará en la superposición de 1s y 0s, lo cual genera distorsión y errores en decodificación; este valor se caracteriza por ser negativo y por ser uno de los parámetros más importantes, ya que permite contrarrestar la dispersión total acumulada en el enlace de la fibra convencional; • la figura de merito define la relación que existe entre la dispersión y la atenuación de la DCF; y • el área efectiva es la sección del núcleo de la fibra por la cual atraviesa toda la potencia lumínica (Gaxiola, 2005).Existen tres esquemas básicos de compensar la CD: • esquema pre-compensación, en donde se compensa la CD antes del tramo de la fibra de trasmisión; • esquema post-compensación, en donde se compensa la CD después de la fibra; y • esquema de compensación simétrica, en el cual se emplean tramos de DCF, antes y después de las fibras de transmisión.La Figura 3 plasma los esquemas planteados.In the configuration in which Raman is injected into the fiber, both EDFA amplifiers are suppressed which allows amplification in any band and lower cost, since the number of devices implemented in the network is reduced.However, Raman amplifiers due to the requiring of high pumping power, add significant cost to the link.The pumping light is injected either in the same direction (co-propagation) or in the opposite direction (counter-propagation) of the input signal.Figure 5 shows above-mentioned configuration.
Figure 6 shows a configuration in which not only Raman is injected into the SSMF-28 fiber through a high pumping power, generally on the order of watt, but also En el desarrollo del presente trabajo se tiene en cuenta la técnica de post-compensación, con el fin de reducir la BER (Bit Error Rate) y presenciar más significativamente los efectos no lineales, debido a un menor decremento del nivel de potencia en comparación con la técnica de pre-compensación (Hoyos & Vélez, 2013;Criollo & Lasso, 2014).De igual manera, la impulsión que debe tener la señal para ser amplificada mediante Raman, requiere que no existan empalmes que generen pérdidas considerables entre el transmisor y la fibra óptica que va a ser usada como medio de amplificación.
En la configuración en la cual se inyecta Raman a la fibra Se suprimen ambos amplificadores tipo EDFA, lo cual permite la amplificación en cualquier banda y menor costo, ya que la cantidad de dispositivos implementados en la red se reduce.Sin embargo, los amplificadores Raman, al requerir altas potencias de bombeo, adicionan costo significativo al enlace.La luz de bombeo se inyecta, ya sea en la misma DWDM Networks When Using Raman Configurations with DCF Fibers.Sistemas & Telemática,15(41), 27-43.an EDFA amplification stage is used.This scheme is called "Hybrid scheme or assisted Raman" and is used in systems of long distances, where the use of DCF is essential.
The pumping required by Raman amplifiers is that of counter-propagation, which is preferred in the design and implementation of distributed Raman amplifier systems (Agrawal, 2002), due to the lower pumping-induced noise coupling on the signal, as well as a less dependence of the gain with the polarization in comparison with the schemes of co-propagation (Neves, Freitas, Almeida, & Calmon, 2003).In fact, the counter-propagation pumping scheme is often the most used, because the transmission signal is less affected by nonlinear effects such as FWM.This is because the distributed Raman amplifier with counter-propagation configuration amplifies the signals of lower power near the end of the span (Neves & Calmon, 2005).Also, the counter-propagation scheme provides a high level of output gain, with the consequence of a higher noise, in relation to other pumping configurations (Grüner-Nielsen et al., 2006).

VI. Optical degradations
Optical degradations are factors that influence unwanted modifications of the original signal, greatly affecting the performance of optical communications systems.Optical degradations are grouped into two categories: linear effects and nonlinear effects, as follows: • linear effects are optical degradations that do not depend on the transmission, Figure 7 shows the classification of linear effects present in the transmissions on optical fiber; • nonlinear effects occur mainly due to high power and changes in the refractive index of the medium, with reference to optical intensity and dispersion
In general, nonlinear effects cause deterioration in the quality of the transmitted signal and force to limit transmission power, number of channels, link distance and transmission rate, among others.The impact of each one of these effects on the signal does not have the same influence (Singh & Singh, 2007).

VII. Optical monitoring parameters
Optical Performance Monitoring [OPM] refers to the physical level monitoring of signal quality, to obtain a good performance of this one in the optical domain (García, 2006).For the analysis and performance evaluation of the optical signal of an optical fiber transmission link, there are a number of parameters and techniques that provide information about the network performance (ITU-T Rec.G.697-2012).These parameters are generally described as follows:

Q factor
It is defined as the Electrical Signal to Noise Ratio [ESNR] at the input of a decision circuit of a receiver; after making a comparison of electrical signals, said circuit allows measuring the quality that has had the link to differentiate the "1" from the "0".A higher value of Q means that the logical levels transmitted are more clearly recognized.The Q factor is considered as a qualitative indicator of the real BER, as it measures both the top and the bottom of the "eye" in order to know the quality of the signal (ITU-T-Rec.G .Sup.39, 2012).

Bit Error Rate [BER]
It is the most important measurement parameter in optical communications for the measurement of system performance; represents the number of bits that have been transmitted del medio, con referencia a la intensidad óptica y a los fenómenos de dispersión (Chan, 2010;Agredo, López, Toledo, & Ordoñez, 2011).En la Figura 8 se muestra la clasificación de los efectos no lineales presentes en las transmisiones sobre fibra óptica.En general, los efectos no lineales provocan un deterioro en la calidad de la señal transmitida y obligan a limitar la potencia de transmisión, el número de canales, la distancia del enlace y la tasa de transmisión, entre otros.El impacto que genera cada uno de estos efectos en la señal no tiene la misma influencia (Singh & Singh, 2007).

Diagrama del ojo
Permite el análisis de las formas de onda de los pulsos que se propagan en el canal de transmisión, de él se pueden deducir parámetros de medición como la BER y el Factor Q. Representa la superposición de las distintas combinaciones de unos y ceros en un rango de tiempo determinado.Además, permite observar parámetros que determinan la calidad de la señal, a través de dos tipos de cruces: • cruce de tiempo, donde se produce la apertura y cierre del ojo; y • cruce de amplitud, el cual consiste en el nivel de voltaje que produce la apertura y cierre del ojo, y se define en un periodo de bit (Dinamarca, 2002;Semiconductor Components Industries, 2014).
Emori, Akasaka, y Namiki (1998) demuestran que la ganancia se satura a medida que se incrementa la potencia de entrada, la saturación casi no tiene dependencia de la longitud de onda, lo cual permite conocer una propiedad importante de las DCF para caracterizar la red; adicional a ello, se mencionan parámetros generales del sistema como canales, λs típicas y una distancia puntual.A esto se suma la revisión de una red WDM, usando polarización erroneously with respect to the total transmitted, measured in reception.BER is affected by noise, dispersion and nonlinear effects; such degradations can be counteracted by the use of dispersion compensators, using amplifiers, increasing transmission power and reducing losses, among other parameters and elements that contribute to maintaining a BER target (Hoyos & Vélez, 2013).

Eye diagram
It allows the analysis of the waveforms of the pulses that are propagated in the transmission channel, from which it is possible to deduce measurement parameters like the BER and the Q factor.It represents the superposition of the different combinations of ones and zeros in a given time range.In addition, it allows the observation of parameters that determine the quality of the signal, through two types of crosses: • crossing time, where the opening and closing of the eye occurs; and • crossing amplitude, which consists of the voltage level that causes the opening and closing of the eye, and is defined in a BIT period (Denmark, 2002;Semiconductor Components Industries, 2014).

VIII. Theoretical approaches and related studies
To date, multiple studies have been made to counteract the limitations of transmission speed and length of the optical links, caused by the attenuation and the different dispersions; in addition, the use of DCFs as an optimal technique to compensate the chromatic dispersion which becomes a degrading alteration of the link.Having said that, it is relevant to have theoretical contributions, coming from advances, research and work related to the topic.In the present study, these are mentioned in their respective chronological order.Emori, Akasaka, and Namiki (1998) show that the gain is saturated as input power increases, saturation has almost no dependence on the wavelength, which allows us to know an important property of DCF to characterize the network; in addition to this, general system parameters such as channels, typical λs and a specific distance are mentioned.Also, revision of a WDM network is added, using combined polarization for four channels; the losses are analyzed in the presence and absence of DCF, evaluating system performance through the noise figure by varying the wavelength.Liao and Agrawal (1999) demonstrate that through numerical simulations, using Raman amplification with a bidirectional pumping scheme and performing dispersion com-pensation, a speed of 40 Gbps is reached in a single channel over transoceanic distances, maintaining a space of 100 km between pumping stations.
For their part, Knudsen and Nielsen (2000) indicate typical dispersion properties for different dispersion fibers, because ultra-high capacity optical transmission and dispersion management systems become unavoidable; this, since their absence of control, will limit significantly the distance of transmission and its correct operation.Considering the dispersion effect previously mentioned, it is necessary to reference Peucheret, Tokle, Knudsen, and Rasmuss (2001), who develop a WDM simulation for eight channels, using new types of dispersion compensating fibers for standard single-mode fibers, in order to evaluate the performance against nonlinear effects and the cross-channel, verifying the attenuation with respect to variations in distance; said study, incorporates conceptual and theoretical references regarding linear effects and dispersion compensating fibers.
As far as the amplifiers are concerned, Pucm, Chbat, Henrie, and Weaver (2001) expose a long-distance WDM transmission scheme for the S-band using dispersion compensating Raman amplifiers, which was successful; they also consider that DCRAs are a key technology for reliable and cost-effective expansion of optical networks in the S-band.
As references concerning DCF, it is possible to consider the study addressed by Cani, Freitas, Almeida, & Calmon (2003), who argue and explain that the SMF-DSF system (100 Km) generates too much noise due to the length of the fiber, but has greater ease in its configuration.Meanwhile, Kimsas, Staubo, Bjornstad, and Slagsvold (2004) analyzed the principle of co-propagation, which consists of inserting insulators between the sections of the DCF amplifier, in order to block the effect of the delayed copy caused by the RBS of the compensating fiber, thus achieving a considerable reduction of losses in the final receiver.Grüner-Nielsen et al., (2006) argue and explain that the reduction of losses is significantly important through DCF and Raman configurations, which is reflected when analyzing the study of a transcendental parameter as property to choose a good DCF, as is the Double Rayleigh Backscattering [DRB] that was evaluated for four types of fibers.This is an undesirable effect for the optical link, in this way was ratified that when DCF is used in the fibers, a better performance is achieved, whereas without the using of DCF the performance was low; however, the work is focused on finding the most suitable properties to choose an optimized DCF.combinada para cuatro canales; las pérdidas son analizadas en presencia y ausencia de DCF, evaluando desempeño del sistema mediante de la figura de ruido variando la longitud de onda.
The exploration work made by Saito, Freitas, and Matos (2007) shows that the increase in Raman gain increases the efficiency for short wavelengths, and also compensates the increase in attenuation, for a fiber of 6 Km in the O-band.Complementing such notions, Shtyrina, Turitsyn, Fedoruk, and Sha (2007) show how system performance can be improved by the pre-encoding source; in addition, the statistics for 3 bits are obtained and it is proceed to quantify the strong dependent pattern ISI.Both studies are based on theoretical references and approaches that may be useful to be referenced and incorporated for the development of the objective proposed in this article.
Returning to some precepts mentioned before, Galdámez (2007) develops a work from the types of existing amplifiers; the focus of his project was based on the two most used amplifiers for the design of an optical communication system in DWDM; these are: the Erbium Doped Fibre Amplifier [EDFA] and the Raman amplifier.Additionally, it corroborates the idea that optical amplifiers are elements that compensate losses and attenuation, to achieve greater distances, amplifying only the photon flow; these are used in WDM and DWDM systems.One of the most suitable amplifiers to use is the Raman amplifier, since it provides characteristics that represent advantages over the EDFA.Such characteristics refer to: that the same fiber is converted into an amplifying medium; having a rather flat gain region, whereby all signal lengths are amplified; and to amplify other bands, in addition to C-band.This project contributes in a decisive way for the development of the proposed objective in this article, due to the incorporation of a significant proportion of the concepts addressed to be carried out.
On the other hand, Forzati, Berntson, and Mºartensson (2008) demonstrate that asynchronous phase modulation increases transmission for the NRZ-OOK system in ultra-long distance WDM systems; in the same line of analysis, we find Alzate and Cárdenas (2011), who highlight advanced modulation formats for the transmission of 40/100Gb/s, compared to the standard technology of 10Gb/s, on optical fiber in short and long distance, as well as some degradations caused by: noise produced by optical amplifiers, chromatic dispersion, PMD and effects of fiber nonlinearities.They also ratify that one of the most widely used options for long-range transmission is Wavelength Division Multiplexing [WDM].
Simultaneously, Orellana (2012) argues that, due to the growing demand for data transport capacity in the metropolitan environment, as a result of the introduction of services and applications with high bandwidth consumption, there indeseable para el enlace óptico, es de este modo, que se revalidó que cuando se utiliza DCF en las fibras, se logra un mejor rendimiento, mientras que sin la DCF el rendimiento era bajo; sin embargo, el trabajo está enfocado en encontrar las propiedades más adecuadas para elegir una DCF optimizada.
For the first time DCF was used as a Raman gain medium because DCFs have high levels of germanium doping and small effective areas, so Raman gains are quite high (Hansen, Veselka, Jacobovitz, & Gruner, 1998), which is why Raman amplifiers currently integrate dispersion compensating and Raman amplification.Several experimental studies of dispersion of the standard fiber and DCF are carried out, integrating the use of DCF in conjunction with different pumping configurations to evaluate the Raman amplification (Brueckner, Ji-hong, & Schuste, 2010).Benavides (2014) works from a specific case, in this sense, characterizes an optical fiber network using DWDM technology from the city of Cotacachi to the sector of Apuela (Ecuador); as a result, he states that the use of optical amplifiers allows amplification of all wavelengths, while increasing signal strength after multiplexing or before demultiplexing.In addition, he emphasizes that the use of Raman amplifiers guarantees a better distribution of power along the optical fiber, and reduce nonlinear effects; Raman amplifiers are used together with EDFA amplification to cover a wider range of wavelengths not covered by the EDFAs.Rojas (2015) delves into wavelength division modulation and dense wave division modulation; he reiterates that DWDM limits the binary speed, due to the linear effect by polarization mode dispersion (PMD), reaction originated by the use of several wavelengths, product of nonlinear distortions due to the displacement and dispersion of the optical fiber.Bonilla (2017), in his study of the hybrid EDFA-RA-MAN amplifier in C and L bands, shows the different combinations of optical amplifiers used nowadays.Raman amplifiers are typically used in co-propagation configuration where the pumping introduced to the optical fiber is performed in the opposite direction to the DWDM signals; therefore, it is ensured that most of the amplification occurs at the end of the fiber path, where signal levels are lower, preventing that the power level at the beginning of each amplification exceed the threshold of nonlinear effects; they operate in both C-band and L-band (Lara & Reis, 2010), but because of their low gain in C-band, the-se amplifiers are commonly used in L-band.One disadvantage to consider when using Raman-EDFA amplifiers is the need for several pumping wavelengths to generate a high flat gain (Bonilla, 2017;Iturri, 2014).
On the other hand, as the number of pumping lasers increases, the waving decreases significantly, although the overall system gain increases by 5 dB for each laser.It is recalled that the use of many pumping lasers is not efficient, but this can be mitigated to a certain extent by increasing the number of channels, using a Raman-EDFA configuration (Martini, Castellani, Pontes, Ribeiro, & Kalinowski, 2009;Martini, Castellani, Pontes, Ribeiro, & Kalinowski, 2009;y Bonilla, 2017).

IX. Conclusions
The bibliographical review shows the multiple experiments and works performed using Raman amplification for transmission in large distance links using DWDM systems.
The increase in transmission power greatly influences the appearance of linear and nonlinear effects that degrade to a great extent the performance of the network, which is why, employing Raman amplifiers, ensures a better distribution of power and, simultaneously, compensates for dispersion and attenuation along the optical fiber.
Due to the great need for data transmission, it was necessary to replace WDM systems with DWDM systems, which allow multiplexing N number of channels or wavelengths by a single optical medium; currently transmit at speeds of up to 40 Gbps.DWDM networks are characterized, as well as the outstanding performance of DCFs to compensate for chromatic dispersion and, in conjunction with Raman amplifiers, to amplify all wavelengths, thus guaranteeing a better power distribution along the optical fiber.
Raman amplification is very useful because amplifies the signal in a greater number of bands, compared to the traditional EDFA amplifiers; however, precaution should be taken with the power level required by the pumping source, since a considerable increase, increments the general level of intensity, which leads to a greater influence of nonlinear effects that degrade the performance of the network.la banda C, estos amplificadores se usan comúnmente en banda L. Un inconveniente a tener en cuenta al emplear amplificadores Raman-EDFA es la necesidad de varias longitudes de onda de bombeo para generar una ganancia alta y plana (Bonilla, 2017;Iturri, 2014).
El incremento de la potencia de transmisión incide considerablemente en la aparición de efectos lineales y no lineales que degradan en gran medida el desempeño de la red, razón por la cual, emplear amplificadores Raman garantiza una mejor distribución de la potencia y, al mismo tiempo, compensa la dispersión y la atenuación a lo largo de la fibra óptica.