RSoft_V11N1_2

RSoft Design UK, Ltd.11 Swinborne Drive,Springwood Industrial Estate,Braintree, Essex CM7 2YP

PHONE: +44 (0)1376 528556E-MAIL: [email protected]: www.rsoftdesign.co.uk

UNITED STATES Corporate headquarters

JAPANoffi ce location

UNITED KINGDOMoffi ce location

JANUARY - JUNE 2012VOLUME 11 NUMBER 1

RSoft: A brief history of your career, company and possible professional associations.I graduated from Montana State University in 2009. My research there focused on the physical layer modeling of soliton-based and non-soliton all-optical WDM systems. I also worked on the analysis and comparison of various modulation formats used for transmitting the control and management signals through a secondary channel created in the fi ber using Electro-optic transducers in Passive Optical Networks. I have considerable industrial and academic experience in modeling fi ber optic systems and networks, and have presented my work at some of the leading technical conferences in the fi eld of Optics. I’m also a member of IEEE. Currently, I am pursuing my Ph.D. at the Georgia Institute of Technology (Atlanta, USA), where I’m working in the Ultra-Fast Optical Communications Lab under Dr. Stephen E.Ralph who also heads the 100G Optical Networking Consortium.

RSoft: Please tell us about 100G Optical Networking Consortium.Sriharsha: The 100G Optical Networking Consortium is an industry-led communications and information technology consortium, createdhere at Georgia Tech. The consortium combines the complementary strengths of the industrial partners and faculty members to enable innovation in and advance the quantitative understanding of optical, electronic and signaling interactions in very high-speed optical networks. The consortium and facility together allow us to investigate a wide range of issues spanning fundamental channel capacity issues related to 100G and 1 Terabit optical transmission to the development of design rules for dynamically reconfi gurable 100Gbps networks. The consortium also investigates optical and electronic technologies that will be used in next generation high-speed optical networks including demodulation algorithms for coherent detection and performance optimization in networks comprised of a wide range of fi ber types. Also, the 100G consortium is now known as the Terabit Optical Networking Consortium.

RSoft: Can you tell us about your current research?Sriharsha: I’m currently investigating different signal processing techniques to improve the signal quality by compensating for the signal degrading effects like crosstalk and modal dispersion in 25G multimode fi ber systems. In addition, our group is developing effi cient demodula-tion algorithms that mitigate the effects of nonlinearities in polarization multiplexed QPSK and QAM-based long-haul WDM systems. We are also investigating other high-speed and high spectral effi cient modulation formats.

RSoft: What kinds of challenges do you face in your research?Sriharsha: Coherent receiver design is a complex subject. The digital signal processing (DSP) has to be fast and effi cient. Dispersion, nonlinear-ities and polarization effects create a number of challenges, often dynamic, in symbol synchronization, channel equalization, etc. It is important to study performance penalties due to each of these individual effects in isolation, which is very diffi cult, if not impossible to do experimentally. For instance, to study the effect of dispersion alone, one cannot simply turn off the noise and nonlinearity effects in real optical fi ber. At the same time, modeling results must be validated by the overall performance from an experimental set up in the laboratory.

RSoft: How has RSoft Design Software assisted in this effort?Sriharsha: OptSim permits modeling granularity at various levels. OptSim readily interfaces with Matlab allowing me to develop and validate standard as well as proprietary algorithms. Furthermore, since the channel and noise statistics are not necessarily stochastic, an actual BER counter model in OptSim was very helpful in pre-FEC BER counting. Also, the strong post-processing features in OptSim allow us to save and export measured data for further processing in external tools. This feature has helped reduce considerable amount of time and effort in my research work. With OptSim, we routinely test innovative ideas that may involve complicated architectures or demodulation methods which are challenging to realize in the laboratory environment due to cost and complexity and time constraints. Lastly, one of our successes has been the excellent correlation between experimental results and simulation results. We are able to do this by carefully including the measured performance of every component and fi ber directly into the OptSim environment.

RSoft: Why did you decide to work with RSoft Design Group for design and simulation software?Sriharsha: Available paths for modeling optical networks include using commercial tools, free tools or writing one’s own custom code. Using industry standard commercial tools permit one to devote maximum time and efforts in actual research rather than on maintaining free or custom codes. The consortium evaluated a number of commercial options and chose OptSim for its versatility, ability to add our own custom Matlab code and for the support we have experienced from RSoft. Also, the GUI is intuitive, and the plotting tools are very good. Lastly, in addition to integrating easily with Matlab, OptSim also works well with other tools like SPICE and BeamPROP.

RSoft: How do you see the need and demand for photonic modeling software in the next generation applications?Sriharsha: Let me comment in light of my current modeling tasks. It is evident that the future installations will focus on coherent (say, PM-QPSK or OFDM) communications. These systems will have to co-exist, for a foreseeable future, with the legacy IM/DD systems. There’s not any other cost-effective way than to use a commercial modeling tool, like OptSim, to estimate performance of such mixed deployments. Furthermore, all next generation optical communications systems will rely more on intense signal processing strategies and photonic modeling software must integrate seamlessly with system developers’ own DSP code.

RSoft: Thank you. We wish you the best of luck in your research.

with Sriharsha Kota Pavan

RSoft Design Group, Inc.400 Executive Boulevard,Suite. 100,Ossining, NY 10562, USA

PHONE: 1.914.923.2164E-MAIL: [email protected]: www.rsoftdesign.com

RSoft Design Group Japan KKMatsura Building 2F,1-9-6 Shiba Minato-ku,Tokyo, 105-0014 Japan

PHONE: +81.3.5484.6670E-MAIL: [email protected]: www.rsoftdesign.co.jp

node pair. For both grooming scenarios, the WDM transmis-sion network contains the same number of fi bers, and the same number of amplifi ers; therefore, the power consumptions due to WDM Lines are equal. However, in End-to-End grooming there are 21 channels that require regenerators, leading to a 2.1 KW increase in power consumption. During the second year the traffi c between each node pair is increased to 15 Gbps. This increase in traffi c lead to more O-E-O switching in the case of Link-by-Link grooming and more end to end wavelengths in the case of End-to-End grooming, increasing power consump-tion due to core routers in both cases. However, the increased traffi c was not suffi cient to demand additional fi bers and amplifi ers; hence, the power consumption due to WDM Lines remained the same. In the End-to-End grooming case, as the number of wavelengths increased, the number of wavelengths that required regeneration also increased. This resulted in increased power consumption due to regenerators. During years 3, 4 and 5 traffi c increase resulted in more fi ber and amplifi ers, so power consumption due to WDM Lines also increased along with other power consuming components. For this example network, as shown in Fig. 2, the End-to-End grooming case consumes less power than the Link-by-Link grooming case.

An accurate estimate of network power consumption is impor-tant to manage the operational expenditure of the network. Estimating network power consumption for optical network is a complex task since many different topologies, grooming methods and optical components exists in a network. MetroWAND can help network planners to model and simulate power consumption in All Optical Networks (AON).

Interview

Sriharsha Kota Pavan

Energy Effi cient All Optical Network (AON) Design.

continued from third page

References

[1] C. Lange, D. Kosiankowski, C. Gerlach, F. Westphal, and A. Gladisch, “Energy Consumption of Telecommunication Networks”, ECOC, 35th European Conference for Optical Communication, Vienna (Austria): 2009, pp. 1-13[2] W. Heddeghem, M.Groote, W.Vereecken, D.Colle, M.Pickavet and P.Demeester, “Energy-Effi ciencyin Telecommunications Networks: Link-by-Link versus End-to-End Grooming”, Conference on Optical Network Design and Modeling (ONDM), 14th Proceedings, Kyoto (Japan), 2010[3] R.S Tucker,”Modeling Energy Consumption in IP networks”,Website:” http://www.cisco.com/web/about/ac50/ac207/crc_new/events/assets/cgrs_energy_consumption_ip.pdf”

rsoftdesign.com

Fiber Optics and Optical Communication group has been established in Delhi Technological University (Formerly Delhi College of Engineering), Delhi in 1999 with the joining of Prof. R. K. Sinha in Applied Physics Department. Initial research mainly focused on the theoretical studies of photonic crystal � bers and electron waveguides. In 2004, Delhi College of Engineering was selected for establishing an advanced R&D centre in the area of Fiber Optics and Optical Communication supported by TIFAC/DST-Govt. of India, Govt. of NCT of Delhi and partners from industries under Mission Reach program of Technology Vision -2020. � is center is known as TIFAC-Center of Relevance and Excellence (CORE) in Fiber Optics and Optical Communication at DCE (now DTU), Delhi.

Under this special initiative program, specialized courses on Fiber Optics and Optical Communication at B.E./B.Tech and M.Sc/M.Tech level with adequate laboratory facilities were started. � is was followed by starting an interdisciplinary M.Tech. programme in Microwave and Optical Communication Engineering jointly with departments of Applied Physics & Electronics and Communication Engineering. Besides this, Optics and Photonics related courses are also o� ered for the students of B.Tech (Engineering Physics) and M.Tech (Nano Science and Technology), where simulation and design of nano scale optical devices are being carried out using RSoft tools.

Doctoral (Ph.D) level research work in the area of Fiber Optics, Optoelectronics, Photonics and Optical communication systems and networks with emphasis on experimental and simulation work is carried out rigorously by over a dozen Ph.D. students in supervision of faculty members of Applied Physics Department of DTU. Research and development activities of this research group mainly focuses on; � eo-retical and Experimental studies of specialty optical � bers and integrated optical waveguides, Nano-Photonics and Multiple accesses techniques in Optical Fiber Communication systems and networks, Design and Development of optical � ber sensors and Opto - transcevier, Development of numerical techniques of light wave and electron wave propagation, � eory and experiments on photonic crystal � bers and on Photonic Bandgap Devices. In the recent past, this group has been actively involved in the design and development of photonic crystal architecture for Slow Light, Negative Refraction and Meta-materials based optical devices using RSoft tool. Optical systems for measurements of high electrical voltage and current, devel-opment of educational kits related to optical � ber communication systems also form an active R&D initiative of this group.

� e group uses various scienti� c software for the design of di� erent compo-nents and their implemen-tation to optical devices and optical communica-tion systems. According to Prof. R. K. Sinha, RSoft software provides the great deal of help in designing and analyzing the group’s research ideas, which have resulted in creation of new knowledge leading research publications in the area of Fiber Optics and Optical communication systems in the leading journals and conference proceedings. Indeed this research group is probably one of the earliest user of RSoft product for the scienti� c research in the area of Photonic Crystal Fiber and Waveguides in India.

Ph. D Students; R&D program at Delhi Technological University.

Figure 2: Evolution of power consumption over 5 years for both Link-by-Link grooming and End-to-End Grooming (LBL:Link-by-Link and ETE:End-to-End).

While optical communication systems are typically digital, a growing body of research has studied applications in the domain of microwave photonics, including the generation, transmission, and processing of optical microwave signals [1]. OptSim, RSoft’s award-winning optical system-simulation package, is ideally suited for the simulation of these technologies and comes with a number of application notes that study microwave photonic systems, including single- and multi-tone RF signal transmission, and both DPSK- and DQPSK-based radio-over-fiber communication.

As an example of OptSim’s ability to simulate advanced microwave photonic applications, we present here the study of two microwave photonic link architectures with balanced detection at the receiver [2]. The first case is based on intensity-modulation with direct detection (IMDD); Fig. 1(a) illustrates this design’s OptSim topology in a back-to-back configuration. A pair of RF sources at frequencies f1 and f2 modulates the output of a 1550-nm CW laser via a balanced-bridge Mach-Zehnder modulator (MZM) with a quadrature bias voltage of Vπ/2. At the receiver, a pair of balanced detectors detects the transmitted microwave signals.

The second case uses a suppressed-carrier (SC) link with coherent heterodyne detection provided by a local oscillator (LO) and a pair of balanced detectors; Fig. 1(b) depicts the OptSim topology of this design, again in a back-to-back configuration. In this case, the balanced-bridge MZM is biased at the null bias-point voltage of Vπ, and a second CW laser acts as the LO whose frequency is offset from that of the source laser by Δf, thereby producing heterodyne detection and allowing for down-conversion of the RF source frequencies. As we shall see, this design demonstrates improved linearity compared to the IMDD case, as well as less susceptibility to distortions due to fiber transmission.

Long-period gratings are widely used as sensors in a variety of fields such as automotive, aerospace, and medicine. Traditional fiber gratings are formed by UV light exposure through a sophisticated phase mask. It has recently been found that a helically twisted fiber can create a similar effect as long-period gratings and can therefore be used as sensors. Light in a helically twisted single-mode fiber will couple into multiple interfering cladding modes. The pitch or grating period of the helix will determine how this coupling happens and therefore how

the device will operate spectrally. Much like other grating devices, helical fibers are phase-sensitive and require the engineer to carefully design the device to fit the desired charac-teristics. RSoft Design Group’s BeamPROP™ simulation tool is ideal for this application as it couples an efficient algorithm suitable for large devices with a robust design tool that allows for arbitrary index profiles.

A helical shape corresponding to the single-mode fiber described in Ref [1] and Ref [2] can be easily created in the RSoft CAD™ and is

illustrated in Fig. 1. A full 3D simulation of the structure was performed using BeamPROP and the results are shown in Fig. 2. Here we show the power in various modes along the structure as well as the total power in the fiber. BeamPROP software makes it easy and convenient for user to make measurements of the field propagating in the device.

BeamPROP’s software package includes MOST™, a utility which automates parametric studies. This utility can be used to study device characteristics as a function of any design parameter. The inset in Fig. 2 illustrates how the grating pitch affects the coupling length. From this, the designer can determine appropriate design parameters and tolerances of a particular device.

A rigorous tool like BeamPROP provides us with more information than a simple analytical model which only accounts for coupling between the fundamental LP01 and the next higher mode LP11 might. BeamPROP, which is based on the Beam Propagation Method (BPM), directly simulates the field propagating in the structure. Therefore all propagating higher order cladding modes are included. Any power not accounted for in the fundamental LP01 and next modes LP11, LP02, and LP12 modes shown resides in even higher order modes. Representative field profiles along the helix shown in Fig. 3 also illustrate the highly-multi mode nature of this device.

For fixed pitch of Λ=0.647mm and grating length L=14mm, we obtain the spectral response shown in Fig. 4 through a parameter scan over wavelength using MOST. Once again this gives the engineer key tool in specifying the design and manufacturing tolerances for this long period grating.

RSoft’s BeamPROP provides a robust approach to study cladding mode coupling in long-period helical fiber gratings.

“Green Networks” is the new buzz word in the telecommunication industry. Energy costs are among the largest operating costs (OPEX) for network operators. More and more network operators are plan-ning to deploy networks that create opportunities for improved energy efficiency, renewable energy usage and carbon emission reduction. Computing the power consumption of both optical access and core networks are of interest to researchers. In [1], Lange et al show the trend of energy consumption in core, aggregation and access networks. In [2], Heddeghem et al describe energy consump-tion for different grooming scenarios in optical backbone networks. RSoft Design Group’s popular optical network planning platform MetroWAND™ now includes capabilities to model and simulate power consumption in All Optical Networks (AON).

In this article, we use MetroWAND to create an example network and compare power consumption in two different traffic grooming scenarios, as discussed in [2]. An Optical backbone network consists of core routers connected by Wavelength Division Multiplexed (WDM) fiber links. A WDM Link can carry multiple optical channels on a single fiber. Each of these optical channels can carry a transmis-sion rate of 2.5 Gbps or 10 Gbps or 40 Gbps. WDM Multiplexers with 40, 80 and 120 wavelengths capacity are available in the market today. In our example network, we use a WDM link capable of carrying 40 channels. Traffic grooming is a process in which sub-wavelength rates are aggregated into an optical channel so that the number of wavelengths used in the network is reduced, hence reducing the capital expenditure (CAPEX) of the network. Using MetroWAND drawing tools, the fiber topology of the National Science Foundation (NSF) network is created as shown in Fig. 1. There are 14 nodes and 20 links in this network. Nodes represent the core router locations and links represent the WDM lines. A WDM line consists of fiber, post-amplifier, line amplifiers and pre-amplifier. Power consumption of the network is the sum of power consump-tions of the core routers and the WDM line equipment. We consider two grooming scenarios: Link-by-Link grooming and End-to-End grooming. In Link-by-Link grooming, all traffic demands are packed into the available wavelengths as efficiently as possible. By so doing, fewer wavelengths, as well as fewer WDM lines, are required to carry the traffic. However, with this arrangement, wavelengths need to be de-multiplexed and unpacked at every node. In End-to End traffic

grooming, wavelengths are dedicated to traffic between a source and destination pair. In this case, all transit traffic are optically by-passed at the intermediate routers so fewer transponder cards are required; however, some wavelengths may not be filled to capacity. Another drawback of this arrangement is that it may require more optical regenerators since the wavelengths are not going through O-E-O conversion at the intermediate nodes.

In this study we considered a planning period of 5 years. A 14X14 traffic demand matrix between each node is created for each year with a growth rate of 5Gbps from previous year. The traffic matrix consists of different demand rates where the smallest granularity is 1Gpbs. The demand rates are groomed into an appropriate num-ber of 10Gbps optical channels. All demands are assumed to be bi-directional. For each traffic demand a route is set up using the shortest path algorithm.

There are two major power consumption components in the net-work: power consumption by the core routers and power consump-tion by WDM lines. Core routers power consumption is plotted using the formula Power [Watts] = Capacity [Mbps]2/3 as proposed in [3]. Power consumption for the WDM lines is mainly due to the ampli-fiers. Amplifiers are placed around 80 km apart and amplifier power consumption is assumed to be 25 W per amplifier. When an optical signal passes through many optical amplification stages, the signal is degraded and needs regeneration. We kept the nominal distance for regeneration at 3000 km and the power consumption for a single channel regenerator at 50 W.

Table 1 shows the results of the power consumption simulation. The first year total power consumptions for both grooming scenarios are almost equal. During the first year, there is less traffic demand between nodes, and traffic bandwidth is equal to 5Gbps between a node pair; so only a few wavelengths are required between each

Important measures of a microwave link’s lin-earity are its higher- order distortion charac-teristics. In the case of a two-tone system such as the designs of Fig. 1, one of the third-order intermodulation products appears at the frequency 2·f1-f2. As we increase the modulation of the RF signals, we expect the detected power of the intermodulation product to initially get worse relative to the detected signal powers and eventually surpass them. As the simulation results in Fig. 2 demonstrate, this is indeed the case. This plot shows the ratio of intermodulation product power to detected signal power at f1, where f1 = 10 GHz, f 2 = 9.5 GHz, Δf = 8 GHz, and Vπ = 5 V. Of particular interest is the point at which the two powers become equal – i.e., the modulation voltage at which the power ratio equals 0 dB. As can be seen, for the IMDD link, this point occurs at an RF modulation voltage of approximately 3 V, whereas in the SC case, it occurs at approximately 6 V. This doubling of the amplitude is consistent with the results in [2] and helps demonstrate the improved linearity of the SC link as compared to its IMDD counterpart.

Another deficit of the IMDD design is its susceptibility to power fading during transmission over optical fiber due to chromatic dispersion [2]. To study this behavior, we simulate both the IMDD and SC designs when transmitting over 30 km of fiber, neglecting the effects of loss and nonlinearities in order to focus on dispersive effects. We transmit a single RF signal over the fiber, with the LO in the SC link set to a frequency offset that produces a down-converted RF signal

frequency at the detec-tor equal to 2 GHz. Fig. 3 shows the normalized detected power as a function of the RF source frequency. As can be seen, in the IMDD link, power fading leads to a strong variation in the detected power as a function of frequency, whereas the detected power in the SC link is very uniform due to the coherent heterodyne detection in the design [2].

As can be seen, OptSim is a powerful tool for studying micro-wave photonic systems, including state-of-the-art designs based on suppressed-carrier modulation and balanced coherent heterodyne detection. Please contact RSoft Design Group ([email protected]) for additional details.

Figure 1: Topology of National Science Foundation (NSF) network.

Table 1: Power consumption of two grooming scenarios over 5 years.

References

[1] G. Shvets, et al, “Polarization properties of chiral fiber gratings,” Journal of Optics, Vol. 11, No. 7, May, 2009.

[2] J. Qian, et al, “ Coupled-Mode Analysis for Chiral Fiber Long-Period Gratings Using Local Mode Approach,” IEEE J. of Quantum Electronics, Vol. 47, No. 11, Nov. 2011.

References

[1] J. Yao, “Microwave photonics,” Journal of Lightwave Technology, vol. 27, no. 3, pp. 314-335, February 1, 2009.[2] C. Middleton and R. DeSalvo, “Improved microwave photonic link performance through optical carrier suppression and balanced coherent heterodyne detection,” Proceedings of SPIE - Enabling Photonics Technologies for Defense, Security, and Aerospace Applications VI, vol. 7700, p. 7700-08, 2010.

Energy Efficient All Optical Network (AON) Design.Simulation of Long-Period Fiber Gratings formed with Helical Fiber.Simulation of State-of-the-Art Microwave Photonic Technologies in OptSim™.

Figure 1: Schematic diagram of a helical fiber.

Figure 1: Topologies for simulating back-to-back multi-tone RF transmission using balanced detection: (a) intensity modulation with direct detection; (b) suppressed-carrier modulation.

Figure 2: Ratio of detected 3rd-order inter-modulation product power and fundamental RF signal power for both IMDD and SC links.

Figure 3: Detected RF power as a function of source frequency for both IMDD and SC links over 30 km of fiber.

Figure 2: Power coupling between LP01 and cladding modes. These results account for mode degeneracy.

Figure 3: Cross-sectional field profiles at different propagation distances.

YEAR

1 2 3 4 5

Link-by-Link Grooming

Core Router(KW) 83.3 170.6 239.1 298.8 353.1

Regenerator(KW) 0 0 0 0 0

Node(KW) 83.3 170.6 239.1 298.8 353.1

WDM Links (KW) 15.4 15.4 18 20.1 25.8

Total Power(KW) 98.7 186 257.1 318.9 378.9

Increase in % 88 38 24 18

End-to-End Grooming

Nodes (KW)

Core Router(KW) 74.7 131.5 177.7 218.5 255.8

Regenerator(KW) 2.3 4.6 6.9 9.2 11.5

Node(KW) 77 136.1 184.6 227.7 267.3

WDM Links (KW) 15.4 15.85 18.8 20.1 27

Total Power(KW) 92.4 151.95 203.4 247.8 294.3


continued on last page

Transverse Field Profile at Z=2mm

X (μm)60- 40- 20- 0 20 40 60

Y (μ

m)

60-

40-

20-

0

20

40

60


X ( μm)60- 40- 20- 0 20 40 60

Y (μ

m)

60-

40-

20-

0

20

40

60

Figure 4: Spectral response of a helical fiber with a pitch of Λ=0.647mm and grating length L=14mm.

x104

Propagation Distance (μm)

0 1 2 3

Mon

itore

d Po

wer

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0Legend:

LP01

LP11

LP12

LP02

TotalPropagation Distance (μm)

13700 13800 13900 14000 14100 14200 14300 14400

Pow

er in

LP 01

mode

0.00

0.01

0.02

0.03 Legend:

Λ=0.645mm

Λ=0.646mm

Λ=0.647mm

Λ=0.648mm

Λ=0.649mm

Wavelength(μm)1.4 1.5 1.6 1.7

Pow

er

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8 Legend:

LP01

LP11

(a)

(b)
























References



References









YEAR

1 2 3 4 5


Core Router(KW) 83.3 170.6 239.1 298.8 353.1


Node(KW) 83.3 170.6 239.1 298.8 353.1

WDM Links (KW) 15.4 15.4 18 20.1 25.8

Total Power(KW) 98.7 186 257.1 318.9 378.9


End-to-End Grooming

Nodes (KW)

Core Router(KW) 74.7 131.5 177.7 218.5 255.8

Regenerator(KW) 2.3 4.6 6.9 9.2 11.5

Node(KW) 77 136.1 184.6 227.7 267.3

WDM Links (KW) 15.4 15.85 18.8 20.1 27

Total Power(KW) 92.4 151.95 203.4 247.8 294.3




X (μm)60- 40- 20- 0 20 40 60

Y (μ

m)

60-

40-

20-

0

20

40

60


X ( μm)60- 40- 20- 0 20 40 60

Y (μ

m)

60-

40-

20-

0

20

40

60


x104


0 1 2 3

Mon

itore

d Po

wer

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0Legend:

LP01

LP11

LP12

LP02


13700 13800 13900 14000 14100 14200 14300 14400

Pow

er in

LP 01

mode

0.00

0.01

0.02

0.03 Legend:

Λ=0.645mm

Λ=0.646mm

Λ=0.647mm

Λ=0.648mm

Λ=0.649mm


Pow

er

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8 Legend:

LP01

LP11

(a)

(b)






















Interview




References


rsoftdesign.com






























References



References









YEAR

1 2 3 4 5


Core Router(KW) 83.3 170.6 239.1 298.8 353.1


Node(KW) 83.3 170.6 239.1 298.8 353.1

WDM Links (KW) 15.4 15.4 18 20.1 25.8

Total Power(KW) 98.7 186 257.1 318.9 378.9


End-to-End Grooming

Nodes (KW)

Core Router(KW) 74.7 131.5 177.7 218.5 255.8

Regenerator(KW) 2.3 4.6 6.9 9.2 11.5

Node(KW) 77 136.1 184.6 227.7 267.3

WDM Links (KW) 15.4 15.85 18.8 20.1 27

Total Power(KW) 92.4 151.95 203.4 247.8 294.3




X (μm)60- 40- 20- 0 20 40 60

Y (μ

m)

60-

40-

20-

0

20

40

60


X ( μm)60- 40- 20- 0 20 40 60

Y (μ

m)

60-

40-

20-

0

20

40

60


x104


0 1 2 3

Mon

itore

d Po

wer

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0Legend:

LP01

LP11

LP12

LP02


13700 13800 13900 14000 14100 14200 14300 14400

Pow

er in

LP 01

mode

0.00

0.01

0.02

0.03 Legend:

Λ=0.645mm

Λ=0.646mm

Λ=0.647mm

Λ=0.648mm

Λ=0.649mm


Pow

er

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8 Legend:

LP01

LP11

(a)

(b)






















Interview




References


rsoftdesign.com







RSoft_V11N1_2

Documents

passive optical networks

terabit optical transmission

g multimode fi ber systems

gbps networks

modeling fi ber optic

rsoft design uk

wide range of fi ber

signal degrading effects