[Vimal, 2(5): May, 2013] ISSN: 2277-9655

IJESRT
INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES & RESEARCH
TECHNOLOGY
IP Over DWDM
Vimal.K. , Dr.T.V.U Kiran Kumar2, G.K.Senthil Kumar3
*1

Department of ECE, Bharath university,Chennai, India

vimal.kailasam@gmail.com
Abstract
This paper completely deals with optical signal transmission; with encapsulation entire raw packets are
converted as optical signals. Since the usage of internet and intranet are increasing day by day, the demands for
bandwidth are also getting increased. In order to meet this demand, DWDM is deployed. DWDM enables service
providers to accommodate consumer demand for ever-increasing amounts of bandwidth. This supports bit rates upto
OC-192.

Keywords: DWDM, IP, Jumbo frame, Optical amplifier, optical multiplexer and demultiplexer.

Introduction provided by both the combination is very high and

Since the internet traffic is growing also it avoids the cost of SONET AND ATM
exponentially, the need for bandwidth was also equipment. There is no need of electrical conversion
getting increased. The concept behind IP over and it supports bit rates upto OC-192 .
DWDM is transmitting data packets through an
optical layer using DWDM. Optical network deals Architecture of IP over DWDM
with end-end delivery, without the need of electrical In telecommunication networks, optical
conversion the data packets are transmitted in optical layer is consider as main transmission layer, the
domain. In optical network, the photonic layer is common transmission layer above the optical layer
responsible to convert raw packets to an optical are SONET, ATM,IP. Types of architecture exist are
signal and the entire management is carried out by closed architecture and open architecture, it is shown
an photonic layer. in figure. 1. The main drawback SONET is suitable
for voice and ATM is suitable for data. The
Need IP over DWDM advantage of using IP is, it supports audio, video,
Initially IP packets cannot be transmitted voice, multimedia etc.
directly through DWDM, the reason is IP packets are
carried through Ethernet frames. Ethernet frames are
electrical frames, this frames are not supported by
DWDM because DWDM fully deals with optical
signal. So it is necessary to provide encapsulation,
then the encapsulated data is mapped onto the
SONET (synchronous optical network) frame i.e.
STM-16 or STM-64. SONET was used as standard
interface between the physical layer and media, the
service provided by the SONET was not very fast and
Figure. 1 Architecture
it offers services at constant bit rate. The major
disadvantage is it cannot be used in bursty internet If we consider closed architecture IP cannot
traffic. In order to reduce the latency and to increase be directly transmitted through DWDM, because IP
the bandwidth, we go for IP over DWDM. data’s are carried by Ethernet frames and it supports
only electrical transmission, and it is not possible to
IP Directly Over DWDM WHY? merge or map this IP data’s directly in SDH frame.
IP supports all data like audio, video, To merge the IP in SDH frame, encapsulation is
images, videoconferencing, multimedia etc the must. Encapsulation is provided by point-point
revenue obtained through IP is very high. DWDM protocol and HDLC framing, and then the
offers ever increasing amount of bandwidth as per the encapsulated frames are directly mapped or merged
customer demand at low cost at low cost. The service
http: // www.ijesrt.com (C) International Journal of Engineering Sciences & Research Technology
[1349-1353]
[Vimal, 2(5): May, 2013] ISSN: 2277-9655

in SDH frame. This originates optical transmission in Here we deal with optical transmission, so it is
the optical layer. necessary to convert all electrical bits into optical
Now consider the open architecture, here signals.
internet protocol data’s are transmitted directly At the transmitter side bits are represented
through DWDM without the need of encapsulation. as 0’s and 1’s, it is a voltage signal. That voltage
The photonic layer is responsible to convert raw signals are used to drive the LASER source, the
packets to optical signal. source is distributed feedback laser, and it is used to
produce optical power as output at some wavelength.
5. Layer functions At the receiver side an photo detector is used to
convert the optical signals into electrical signal.

2. Super Jumbo Frames

Figure.4 is frame format of an jumbo frame.
The preamble consists of seven bytes all of the form
10101010, and is used by the receiver to allow it to
establish bit synchronization, if there is no clocking
information nothing is being sent. The SFD is a
single byte, 10101011, which is a frame flag,
indicating the start of a frame. The source and
destination address are 6 bytes, it contains the address
Figure. 2 Functions of layers of the sender and receiver. The length/type indicates
Functions of layers are shown in figure. 2 the number of data in the frame i.e. payload, the
and it is clear that the application, presentation and maximum payload utilized in jumbo frame is 48
session layer take the input and converts it into data, bytes to 9000 bytes. Frame check sequences are used
then the transport layer adds a segment header for error detection purpose at the receiving end.
converting the data into segments, and the network
layer adds a network header and convert the
segments into packets.
The data link layer adds a frame header
converting frames into bits and the MAC sub layer is
converts the frame into number of bits. The physical
layer is responsible for selecting the suitable medium
in order to transmit the number of bits.

1. Photonic Layer
Photonic layer is responsible for converting Figure. 4 Frame format
raw packets into optical signal, and the entire Normally IP data’s are carried by an
management in an optical network is also carried by Ethernet frame, which has the maximum transmission
this layer. unit as 1500 bytes. That can produce optical output
signal but is not large enough to transmit over a long
distance. The drawback is if any frame gets damaged
or corrupted once again it has to be transmitted, so
the time to process in all layers will be more because
of that performance throughput gets affected. To
avoid all these problems super jumbo frames are used
were the maximum payload is 9000 bytes, the optical
power produced by this frame is large enough to
transmit over a long distance. In case of error, during
retransmission the process time also get reduced,
with this 99.5% throughput can be achieved .
Figure. 3 Photonic layer operations

The above figure. 3 clearly explain clearly 7. DWDM System

about the operation of photonic layer. The physical Dense Wavelength Division Multiplexing
layer is responsible for transmitting bits through the (DWDM) is the process of multiplexing signal of
transmission medium, those bits are electrical bits.
http: // www.ijesrt.com (C) International Journal of Engineering Sciences & Research Technology
[1349-1353]
[Vimal, 2(5): May, 2013] ISSN: 2277-9655

different wavelength onto a single fiber. Through this

operation, it creates many virtual fibers each capable
of carrying a different signal. At its simplest, DWDM
system can be viewed as a parallel set of optical
channels, each using a slightly different light
wavelength, but all sharing a single transmission
medium. This new technical solution can increase the
capacity of existing networks without the need for Figure. 6 (a) Audio Signal
expensive re-cabling and can reduce the cost of
network upgrades.

Figure. 6 (b) Video Signal

Figure. 5 DWDM Systems

DWDM performed on an optical fiber serves Figure. 6 (c) image Signal

as the underlying carrier for the optical network. The The source, a solid-state laser, must provide
narrow channel spacing of about 1nm characterizes stable light within a specific, narrow bandwidth that
the system. Fig.1 depicts the general structure of the carries the digital data, modulated as an analog
DWDM system. The Erbium doped Fiber Amplifier signal. Here we have taken audio, video, image
(EDFA), Multiplexer and the Demultiplexer form the for transmission through DWDM, depend upon the
vital blocks of the system. capacity of the data optical signals are produced.
The concepts of optical fibre transmission, Figure 6(a) 6(b) 6(c) shows the optical signal
loss control, packet switching, network topology and representation of audio, video and image data.
synchronization play a major role in deciding the b) Multiplexer
throughput of the network. The transmitters are wavelength dependent,
so each and every transmitter should transmit their
Simulation signals in its specific wavelengths. The function
Here we consider three types of data video, performed by the multiplexer is to combine all the
audio and images, each are transmitted at different signals at different wavelength and are transmit
wavelength. Sine wave is used to represent the through the single fiber. The simulation related to
optical signals, which are shown below. Simulation multiplexer was shown in figure. 7.
was performed at the physical layer by using .NET.
This was explained step by step with the help of
output produced in every stage.
a) Generating the signal
The IP packets are carried by Ethernet
frames, type of Ethernet frame we preferred is super
jumbo frames which has the maximum transfer unit
9000 bytes. At the physical layer data are represented
as bits i.e. voltage signals. This voltage signals drives
the tuneable transmitter and produce the optical
power as output.
Figure. 7 Multiplexing the signals

Some inherent losses occur in the

multiplexing block; the losses depend upon number
of transmitters or no of users. Those losses can be
http: // www.ijesrt.com (C) International Journal of Engineering Sciences & Research Technology
[1349-1353]
[Vimal, 2(5): May, 2013] ISSN: 2277-9655

controlled by varying parameters like optical power

and channel spacing. The above figure. 7 shows, the
optical signal produced by all transmitters are in
phase with each other.
c) Transmission Media
The transmission media is needed to
transmit the signals from one host to another host, the
preferred transmission media is single mode
fiber,G.652 is commonly used but it is characterized
by large amount of chromatic dispersion at C (1530- Figure.9 Optical amplification
1565 nm) and L (1565-1625 nm) bands, needs The output of Optical amplification is
dispersion compensation when a system has a shown in figure. 9 The output of optical
transmission rate exceeding 2.5 Gbps. When G.652 amplification is shown in figure. Since we deal with
fiber is operating at 10 Gbps, dispersion optical network, inline amplifier is used for
compensation is costly so it is not used for long amplification along the transmission channel. The
distance communications. So we go for dispersion function performed by optical amplifier is to boost
shifted fiber. the incoming signal in order to reduce compensate
G.653 fiber is a kind of dispersion shifted the losses which occur in the signal during
fiber (DSF), whose dispersion is greater than -1 transmission. There are number of optical amplifier
ps/nm/km and less than 3.5 ps/nm/km at C and L available among that the suitable simplifier is erbium
band regions. It has the smallest amount of chromatic doped fiber amplifier. EDFA boost the optical signal,
dispersion at C band. The maximum dispersion value without the need of electrical/optical (E/o) or optical
for G.653 fiber is about one sixth of that for G.652 to electrical (o/E) conversions. EDFA can be used in
fiber. With the zero-dispersion wavelength of G.653 co-propagation direction and also counter
fiber set at 1550 nm band, systems can operate at bit propagation direction it has two pump which operate
rates of 20 Gbps or 40 Gbps. at 980nm and 1480nm. Among two pumps only
980nm is preferred. The reason is noise at this
wavelength is very low compared to 1480nm. The
gain also can be active up to 30dB.
e) Demultiplexer
De-multiplexing is the process of separating
the signals; the output of de-multiplexer is shown in
figure. 10, from the above figure the audio signal,
video signal and image signal are separated and are
received by the particular receiver.
Figure. 8

Therefore, G.653 fiber is the best choice for

ultra long-haul transmission over a single
wavelength. Dispersion-shifted fibers have been
created to take advantage of the best of both worlds:
by shifting the zero-dispersion wavelength into the
1550nm band we also operate at the lowest possible Figure. 10(a) Audio signal
attenuation. Figure .10(a) represents the audio signal, the
d) Amplification audio signal transmitted at the transmitter is received
The losses occurring at the multiplexer are after amplification without any distortion.
compensated and In order to transmit signals over
long distances it is necessary to compensate for
attenuation losses within the fiber.

Figure. 10(b) Video signal

http: // www.ijesrt.com (C) International Journal of Engineering Sciences & Research Technology
[1349-1353]
[Vimal, 2(5): May, 2013] ISSN: 2277-9655

achieved because the retransmission rate, loss of

Figure .10(b) represents the video signal, the frame during transmission was reduced, the super
signal transmitted at particular wavelength is jumbo frame is a four times bigger than the fast
received at the receiver. Ethernet frames. So by increasing the maximum
transmission unit in a frame very high throughput
was achieved.

References
[1] Dong Tianlin, “Fiber-optic communications
and fiberoptic information network”
.Beijing: Tsinghua University Press, [2]
Figure. 10(c) Image signal [2] Zhuang Jianzhong, “DWDM optical
transmitter design” CATV Technology,
Figure .10(a) represents the image signal; No.12, 2006: 77-812005.9.
the multiplexed signals can be separated by using the [3] B.T.Doshi, S.Dravida, P.Harshavardhana,
prism or optical grating. Normally prism are used to M.Qureshi, "Comparison of next-generation
separate the signals along the fiber which propagate IP-centric transport architectures", Bell
at different wavelength can be separated. During de- Labs Technical Journal, Oct-Dec 1998,
multiplexing also some losses will occur those losses pp.63-85
are compensated by using post amplifier; it is placed [4] Kartalopoulos, S. V. Introduction to DWDM
in front of the receiver. The transmitters are Technology: Data in a Rainbow; SPIE
wavelength dependent but the receivers are not Engineering Press: Bellingham, WA,
wavelength dependent. 1999; 256 pp.
[5] O.Gerstel, P.E. Green, R. Ramaswami
Conclusions Architecture for an Optical
The performance analysis of this paper deals Network Layer,Feb 1996
with throughput load and delay, our main objective is
to obtain maximum throughput. Initially we use fast
Ethernet or gigabit Ethernet frame to carry the data,
these frames have the maximum transfer unit as 1500
bytes. The throughput obtained by using this frame is
72% only its not enough so we use super jumbo
frames, it has the maximum transfer unit 9000 bytes,
throughput obtain through this frame is 99.5%.

Figure. 11
From the above figure. 11 the maximum
throughput achieved is ~100% i.e. 99.5% this is

http: // www.ijesrt.com (C) International Journal of Engineering Sciences & Research Technology
[1349-1353]