Simulative Analysis of a WDM RoF Architecture for Full Duplex Wired and Wireless Network THESIS REPORT Submitted in partial fulfillment of the requirements for the award of M.Tech Degree in Electronics and Communication Engineering (Microwave and TV Engg.) of the University of Kerala Submitted by Karthikeya G. S Fourth Semester M.Tech, Microwave and TV Engg. Guided by Pradeep R Assistant Professor Dept. of ECE DEPARTMENT OF ELECTRONICS AND COMMUNICATION ENGINEERING COLLEGE OF ENGINEERING TRIVANDRUM 2012
Optical communication offers tremendous bandwidth, but does not allow mobility. On the other hand wireless technology supports mobility but fails to provide high data rates. Radio over Fiber is an amalgamation of these two technologies. In this thesis work, a novel architecture for full duplex wired and wireless network with optical fiber back- bone network is proposed. We employ a dual drive dual port Mach-Zehnder modulator to simultaneously modulate the CW laser source with wired and wireless data, for the downstream link. Light remodulation technique using RSOA is utilized in the upstream link to provide a reliable bidirectional optical channel. WDMis used to further increase the capacity of the system. Simulation of the proposed scheme demonstrates a 2.5 Gb/s downlink and 1 Gb/s uplink.
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Simulative Analysis of a WDM RoF Architecture
for Full Duplex Wired and Wireless Network
THESIS REPORT
Submitted in partial fulfillment ofthe requirements for the award of M.Tech Degree in
Electronics and Communication Engineering (Microwave and TV Engg.)of the University of Kerala
Submitted by
Karthikeya G. SFourth Semester
M.Tech, Microwave and TV Engg.
Guided byPradeep R
Assistant ProfessorDept. of ECE
DEPARTMENT OF ELECTRONICS AND COMMUNICATION ENGINEERINGCOLLEGE OF ENGINEERING
TRIVANDRUM
2012
DEPARTMENT OF ELECTRONICS AND COMMUNICATION ENGINEERINGCOLLEGE OF ENGINEERING
TRIVANDRUM
CERTIFICATE
This is to certify that this thesis report entitled “Simulative Analysis of a WDM RoFArchitecture for Full Duplex Wired and Wireless Network ” is a bonafide recordof the work done by Karthikeya G. S, under our guidance towards partial fulfillmentof the requirements for the award of Master of Technology Degree in Electronicsand Communication Engineering (Microwave and TV Engg), of the University ofKerala during the year 2012.
Pradeep RAssistant ProfessorDept. of ECECET(Guide)
Dr. Vrinda V NairProfessorDept. of ECECET(Thesis Coordinator)
Dr. Jiji C.V.ProfessorDept. of ECECET(P.G. Coordinator)
Prof. J DavidProfessorDept. of ECECET(Head of the Department)
ACKNOWLEDGEMENTS
I would like to thank my Guides Mr. Pradeep R, Dr. N. Vijayakumar, Thesis Co-
ordinator Dr. Vrinda V Nair, PG Co-ordinator Dr. Jiji C. V, HOD Prof. David and
all others who have supported me for the fulfillment of this work.
I would also like to thank the entrepreneurs who started vegetarian hotels in
Thiruvananthapuram.
Karthikeya G. S
ii
DEDICATION
To my sister Shrilalitha Girish
To the invisble angels described by Richard Feynman in his lectures on Physics
To the magical eyes of senorita
iii
ABSTRACT
Optical communication offers tremendous bandwidth, but does not allow mobility. On
the other hand wireless technology supports mobility but fails to provide high data rates.
Radio over Fiber is an amalgamation of these two technologies. In this thesis work, a
novel architecture for full duplex wired and wireless network with optical fiber back-
bone network is proposed. We employ a dual drive dual port Mach-Zehnder modulator
to simultaneously modulate the CW laser source with wired and wireless data, for the
downstream link. Light remodulation technique using RSOA is utilized in the upstream
link to provide a reliable bidirectional optical channel. WDM is used to further increase
the capacity of the system. Simulation of the proposed scheme demonstrates a 2.5 Gb/s
downlink and 1 Gb/s uplink.
iv
Contents
1 Introduction 1
1.1 Relevance of RoF technology in today’s world . . . . . . . . . . . . 2
3.9 log BER versus received power . . . . . . . . . . . . . . . . . . . . 33
ii
CHAPTER 1
Introduction
Communication is the key strength for man’s progress. It is the power, which, supports
our existence and promote our growth. Communication could be verbal or non-verbal.
As scientists we would desire to invent machines and associated technology capable of
unlimited bandwidth for an unlimited time, gargantuan broadband speeds, HD clarity
video conferencing, Live broadcast of 1080p quality videos from any part of the globe,
3Dimesional video and audio streaming. For all these utopian features of communi-
cation system to evolve one needs to thoroughly comprehend the beautiful technology
underlying, and that is optical communication.
There are millions of reasons for the grand success of the optical communication
technology. Some of them are high bandwidths, high immunity to noise, and easy
maintenance. The major drawback is that it is fixed or stagnant terminals. But wireless
technology on the other hand, provides mobility but falls short in bandwidth issue. “Ra-
dio over fiber ”is the magnificent amalgamation of wireless and optical communication.
Benefits of Radio over Fiber:
• Mobility
• Flexibility of devices
• Highly economical in the long run
• Mobile Broadband internet
• Transmission of millimeter waves
• Immunity to radio interference (in the fiber)
• Reduced power consumption
• Dynamic Resource allocation
Limitations of Radio over Fiber:
• Poor dynamic range
• Device nonlinearities can be detrimental
• Relative Intensity Noise in laser sources
• Dispersion effects in the fiber
1.1 Relevance of RoF technology in today’s world
The world currently has 560 crores of active mobile handsets (80 % of popula-
tion). India has around 88 crores of mobile handsets (73% of the population). The
revenue generated by telecom industries in India alone accounts to Rs.117,039 crore
With a penetration of this magnitude, telecommunications industry has become one of
the most important industries in the country and worldwide.
With the advent of telephone system people were not satisfied with that service so
soon data communication: internet on the go evolved. But initially this supported very
low data rates just sufficient to open the browser and check a couple of emails. Even-
tually, the third generation flourished, which paved way for a technology that mankind
had never heard before.
With 3G people could browse and download data at a rate of 7Mbps. That is equiv-
alent to downloading a 700 MB movie in fewer than 15 mins, that’s lightning speed.
1.2 Need for Hybrid technologies
The world economy is dependent on the thousands of computer networks all around
the globe. Any end user would always like to be “connected ”anytime and anywhere.
That necessitates the researchers and engineers to design networks which are fool-proof
i.e., to provide reliable connection to the broadband subscribers. But the type of ser-
vices is not at all uniform or standardized. In other words, there are myraid of types of
networks and the services available to the subscribers.
2
The subscribers could connect to the internet or avail other broadband services such
as video conferencing or file sharing, in a variety of ways like GPRS (wireless connec-
tion), DSL (wired connection) or FTTH (wired connection). Thus we observe a mixture
of multiple technologies prevailing in the market. The main reason due to lack of com-
mon type of technology to all subscribers is due to lack of rapid up gradation. If the
companies embrace the new and better technology as and when it is invented, then the
chances of single type of connection could be expected after a couple of years.
Hence we need architectures which cater to the need of hybrid technologies. These
networks would be future-proof for at least a decade or so. In this thesis work, the au-
thor concentrates on the amalgamation of wired and wireless technologies.
The thesis report is organized as follows: chapter 2 gives a broad overview of radio
over fiber technologies with special emphasis on full duplex wired and wireless archi-
tectures. chapter 3 illustrates the proposed scheme followed by results and applications
3
CHAPTER 2
Literature Survey
Man has always been interested in exploring novel ways to communicate with fellow
humans from the ancient times. Initially man communicated with gestures and grunts
which could convey the arrival of the prey or other danger. But eventually man realized
the necessity to look at life in a fresh perspective, because his immediate needs for sur-
vival were naturally satisfied. Thus a major time gap grew in human’s life, during this
time he went beyond survival and embraced science.
As years progressed, civilizations were built with utmost care and charm. Gradually
industries sprung up which were specialized in mass production of specific products,
this slowly tempted man to be selfish. Due to this evolution, man grew as a consumer of
goods and information, as gargantuan production of goods demand voluminous infor-
mation about the technology. This naturally involves communication between human
beings, and machines alike. Due to this phenomenal development of communication
revolution, high capacity data transfer between terminals in a network has become more
than a necessity today. Billions of rupees is invested in order to economically establish
data transaction within a nation and overseas alike.
Due to the attractions of multimedia technology, optical fiber communication was
one of the natural choices. Light offers unimaginable data rates (if intelligently de-
signed), this is due to the fact that the frequency of light is hundreds of Tera Hertz. But
this eye-catching feature of optical communication fails miserably when it comes to
mobility of the end terminal, i.e., optical fiber would terminate onto a designated place
in the network. It is a highly demotivating feature of the optical fiber technology.
The communication via radio waves has been a well established technology, but
unfortunately it does not inherently support the high bandwidths like the optical fiber.
Thus instead of acting as complementary technologies wireless and optical communi-
cation supplement each other, which is the essence of radio over fiber. The beauty of
this famous culmination of both the technologies is that, it offers fantastic data rate
accompanied by portability of the terminal. The backbone would be the optical fiber,
buried silently inside mother earth and the last mile connections would be established
by wireless technology. Radio over fiber would ensure that crucial data communication
would definitely occur, provided the subscriber is in the coverage area of the network
operator.
2.1 Origin of the concept
Daniel Colladon demonstrated light guiding in water in 1884, this experiment paved
way for more explorations with light and its interaction with transparent materials. Af-
ter, a couple of years later, illuminated fountains became famous. Thus, light guiding
had humble beginnings in recreation and entertainment, especially in Europe. The idea
of an optical channel is basically “looking ”at a far distance, but the image seen would
be blurred and slightly distorted.
After years of research, engineering and redesign, the concept of optical telegraphs
and photophones were explored in the late 19th century [1]. But serious research of
optical communication as seen today could be traced back to the classic paper [2] in-
vestigated by Charles Kao, “Godfather of Broadband ”, in 1965 for which he received
the Nobel prize in Physics for “groundbreaking achievements concerning the transmis-
sion of light in fibers for optical communication ”in 2009. After the birth of that idea
a lot of researchers especially in USA, Japan and Europe experimented on achieving a
high data rate transmission in the optical fiber. Meanwhile research on millimeter wave
generation was also progressing rapidly. During the 1960s there was an intellectual race
between optical fiber communication and millimeter wave communication [1].
The failure of millimeter wave technology is evident in the incident described be-
low: The British Post Office had laid a 14.2km millimeter waveguide, due to the tem-
perature fluctuations the waveguide would undergo some structural changes, which in-
5
turn would lead to unnecessary mode formations in the waveguide thereby reducing
the transmission capacity tremendously. Thus in order to compensate for the unwanted
modes, the engineers designed an automated machine which compensated for the shape
changes of the waveguide. This design was uneconomical and highly unreliable. After
this design failure researchers never invested their time in millimeter waveguides for
commercial transmission and broadcasting of signals [1].
Initially, inefficient LED in conjunction with multimode fibers was used. But this re-
quired lot of regenerators for a relatively short distance. But it was known that the glass
fibers had low attenuation at specific wavelength ranges for instance around 1.55um.
Thus researchers developed solid state laser sources which operated specifically
near those wavelengths but initially the laser were multimode, though it improved the
existing scenario, did not revolutionize the industry. The breakthrough happened after
the development of single mode laser source such as heterostructures whose linewidth
was extremely narrow. This was used along with the single mode fiber leading to an
increased distance without any regenerators for short haul links. Single mode fibers are
still in use today for long haul and ultra long links (but with necessary repeaters).
The next ground breaking concept was the extension of the frequency division mul-
tiplexing (a well established concept in wireline communications) to optical domain in
1990s, formally called as wavelength division multiplexing, which is basically trans-
mission of different data sources in different colours. This is one of the most successful
inventions in the recent decades. WDM also paved the way for development of optical
amplifiers such as Erbium Doped Fiber Amplifiers (agnostic or colorless amplifiers)
operating in the 1.55um attenuation window. Also, these amplifiers are broadband in
nature thus supporting hundreds of colours in the same fiber. Consequently numerous
devices and custom made devices came into existence to cater the needs of the wave-
length division multiplexing system [3].
Meanwhile, wireless technology was also making significant changes in the market,
industry and economy. Started with transatlantic communication conducted by Mar-
6
coni in 1901, wireless technology created the stir in broadcasting and communication
segments. Radio made it’s entry as early as 1920s with numerous broadcasting stations
worldwide (some of which are active even today). Very soon TV broadcasting also
happened. Bell labs developed cellular concept. Satellites specific to communications
were launched during 1960s. Military needs also promoted much of the development
of wireless technology.
Cellular phone, Wi-Fi, WLAN made their grand entry during the early 1990s. WiMaX
and EDGE technologies also entered the wireless arena during the onset of the millen-
nium. Another popular classification for the mobile era is first generation (1979-1985
or so) which housed only analog communication and voice support only existed. Sec-
ond generation (1990 - ) fully digital system and currently the third generation this
supports multimedia, internet on the go and video calling. 3G systems already had de-
ployed thousands of kilometers of optical fiber exclusively for mobile telephony. There
are over 560 crore mobile handsets operative globally (meaning 79.86% of the popula-
tion). India stands second largest consumer of cellular technology with 88 crore mobile
phones (73.44% of population), of which only a mere 14% is on the 3G network. Thus
there is an unprecedented requirement of research, design and implementation of the
forthcoming high data rate technology, which is basically a more evolved and mature
“Radio over Fiber ”[4], [5]. RoF promises more than 7 Mbps data rate to the end user.
Thus with billions of subscribers globally and the penetration of broadband technol-
ogy would pave the way for enhancements in radio over fiber technology. In this thesis
work the author is keenly interested in reliable transmission, distribution and broadcast-
ing of wireless signals in the optical communication channel. Hence the wireless aspect
of this technology is not addressed. As standard equipment along with reliable design
procedures are successfully ruling the cellular industry[6].
7
2.2 Overview of Radio over Fiber technology
Radio over fiber (RoF) technology was first explored in 1991. The problem they set
out to solve was to provide mobility to business and residential users, wherein the radio
transmission would be carried out using telegraph poles. The coverage was intended to
be only 200-300m for a local area [7].
The problem of microwave photonics became relevant only after huge success of
cellular communications. After the development of solid state lasers in early 1960s,
lot of improvements has happened to the laser sources as such. The linewidth of laser
source is extremely narrow. And customized laser sources have been developed. Laser
arrays are also available for WDM systems. Modulation of solid state lasers by 40 GHz
signals were explored as early as 1996. Direct modulation is relevant in analog CATV
applications, but the input data rate is limited to 10 Gbps. Indirect or external modu-
lation is popular for high speed data transmission, the most common example includes
Mach Zender modulator which supports error-free 40Gbps data transaction.
Receivers also have wide bandwidth with reasonably decent switching speed. Usu-
ally pin photodetectors is preferred over avalanche photodiodes. Also commercially,
single mode fiber with step index with dispersion compensation is used. Normally, the
entire optical communication is centered on 1.55um. The reason being low attenuation
and mature technology of EDFA (Erbium Doped Fiber Amplifiers), which in turn oper-
ates in that wavelength window.[8]
Problems in an analog link (not specific to RoF technology, but in transmission
of microwave signals in light in general) are multifold such as, harmonic and inter-
modulation distortions of the source, nonlinearities in the external modulator, nonlin-
earities in the fiber and all-optical amplifiers, nonlinearities in the photodetectors.
Optical transmission and distribution of wireless signals can be realized as :
(a) RF over fiber, suffers due to enhanced effects of chromatic dispersion.
(b) IF over fiber, increased complexity in hardware realization
8
(c) Baseband over fiber, simplest form of transmission but does not support high data
rates.
Problems faced by optical transmission of mm-wave signals:
Due to the effects of nonlinear modulator, the analog mm-wave is weakly intensity
modulated onto the light carrier. The nonlinear characteristics of the external modula-
tor play a detrimental role in the mitigation of the actual signal strength in light.
Fiber chromatic dispersion places an upper bound for the transmission distance. The
opportunity for researchers is that useful information is low (<3GB/s as reported in [9])
compared to the occupied spectrum 40 GHz. Nonlinear effects at the detector also plays
a key role in deteriorating the signal.
Figure 2.1: Major problems encountered in radio over fiber links
Different modulation strategies for optical mm-wave are as follows:
(a) Intensity modulation: simple to generate but spectrally inefficient and certain
power penalty issues.
(b) Optical single sideband with carrier: spectrally efficient but unsuitable for long haul
communication.
9
(c) Optical carrier suppression technique: spectrally efficient but requires a larger drive
RF power.
The basic architecture for a conventional central office, base station etc. is clearly
discussed in [9]. A generalised basic architecture is illustrated in figure 2.2 [10]. This
scheme is usually implemented in the urban areas with lot of users of different genre.
Each baseband dataset is carefully modulated onto a specific optical carrier and then
central unit (CU) would redistribute these signals onto the respective terminals, wherein
the O-E conversion and further transmission in the electrical domain takes place within
the coverage area.
Figure 2.2: Genralised architecture for distributed antenna systems
10
2.3 Radio over Fiber link design
There are several factors specific to the design of a RoF link which need to be con-
sidered while designing a real-world system. The actual system depends on geography
and the traffic load in that particular geographical location.
(a) Carrier frequency: This would decide the type of optoelectronic components to
be incorporated in the system. Also, higher frequency supports higher bandwidth, but
this would increase dispersion in the fiber.
(b) Radio channel bandwidth: This in turn depends on the carrier frequency and the
allocated number of users. The CNR would decide the allowable bandwidth.
(c) Number of channels: the maximum allowable transmitted noise and stability re-
quirements decide the number of channels for a particular remote station.
(d) Modulation format: as discussed earlier, simple modulation schemes are simple
to design and economical but spectrally inefficient. Thus if the market prediction is
high traffic growth, then complex schemes which can accomodate a lot of users would
prove to be economical in the long run.
(e) Cost: is another important criterion for the deployment of practical systems, it
depends on the short term and long term profits.
(f) Distance: of operation is another parameter to be considered for the length of the
fiber, operative powers of the laser diodes and hence the data rates.
(g) Hybridity: A mixture of wired and wireless users might be important if the op-
tical feeder network is future-proof[10].
Basically, for a given scenario the engineering team decides the predicted traffic
11
load. Then it would be divided among various smaller geographical cells. Then the
link specific parameters are computed and an extra margin is always allocated to ensure
reliability in case of an outage. After this process, the most suitable and economical
components are installed and tested for the designed specs, if they do not match then
the link would be redesigned.
2.4 WDM RoF networks
WDM networks are used for ultra high capacity trunk and metro networks. Different
colours are allocated to different base stations, but with the same optical fiber backbone
network [11]. This scheme is future proof since trouble-shooting of the optical network
is relatively easy. The WDM network could be of various types such as
(a) Point to point links
(b) WDM star networks
(c) WDM ring network
(d) Wavelength interleaved RoF system
(e) A combination of the above.
2.5 Full Duplex Radio over Fiber schemes for multigigabit commu-
nications
Full duplex RoF schemes were reported earlier with baseband modulation but com-
plete bidirectional scheme operational above 60 GHz was carried out in the early 2003
[12], [13]. In the early schemes efficiency was not at all the criterion, exploration of
light as carrier to millimeter waves was the primary agenda and it was successfully
demonstrated with a bit rate of 155.52 Mb/s. The scheme was to modulate the CW
laser source using an EAM at the transmitter and a pin diode was used as the receiver.
This scheme used two individual systems to implement downlink and uplink. Hence-
forth two individual fibers were deployed (23 kms in length). Also it addressed the RoF
link from Central Office to the Base station. This also demonstrates that by conducting
the experiment in that short length would prove that the scheme could be generalized to
any extent [14].
12
Hybrid technologies are equally important for economical deployment of the net-
works. In other words, the already existent hardware should be future proof; even if a
new technology is unraveled, the hardware already laid out should permit slight modifi-
cations to the system to accommodate the change. Sometimes, a dark fiber is laid out for
this purpose. In that context hybrid schemes are very helpful. Here it was demonstrated
that baseband (CATV and other broadcasting applications), microwaves (wireless sig-
nals) and millimeter wave signals could be transmitted over a single channel (optical
fiber) this was accomplished using a multiplexer/ demultiplexer pair at transmitter and
receiver pair. This paper proved that having hybrid technology would be future proof
(assuming a prediction of linear growth of traffic). Also it operated on a high speed data
of 2.5 Gbps [15].
A simple, yet elegant solution is a cascaded semiconductor optical amplifier (SOA)
and Electroabsorption (EAM) modulator which serves the dual purpose of converting
the frequency of operation due to the device nonlinearity and meanwhile performs the
routine job of amplification and modulation. SOA is used for frequency up conversion
whereas EAM is used for frequency down conversion. But this scheme demonstrated
experimentally is highly not recommended for multigigabit communication. On a pos-
itive note suppose the SOA-EAM chip becomes commercially cheap product then this
scheme has a potential for small local loops and LAN applications. But the problem
would be of standardization, which cannot be easily handled [16]. It is interesting to
note that a lot of researchers are always interested in providing solution which contains
devices which perform dual functions.
Fully fledged bidirectional RoF communication systems catering to the needs of
millimeter signals is very challenging task to accomplish. This is due to the inherent
problems existing in the transmission of optical signal carrying data with humungous
rate.
13
Figure 2.3: RoF scheme for SOA-EAM bidirectional RoF system
One of the schemes is to suppress the carrier to produce two side bands and mod-
ulate the data onto one of the sidebands. The optical carrier suppression could be
achieved by LiNbO3 modulator suitably biased in the nonlinear region.
Further a copy of incident light is combined with the optically suppressed and mod-
ulated light to the fiber. At the destination a Fiber Bragg grating is used to reflect the
unmodulated light back to another MZM modulator (which appends the uplink data
onto that light). The remaining light which is suppressed and modulated is used for
detection purpose. Hence a simple solution for the bidirectional RoF is achieved. But
there are problems in this scheme, as it uses two fibers and not suitable for long-haul
applications. Also it is observed that there is quite wastage of bandwidth since the re-
dundant light is transmitted at the Central Station, which eventually dies out at the Base
station thus needs an optical amplifier which is expensive [17]. However, this scheme
gives a brilliant insight into the working of the bidirectional RoF link.
Having said that devices serving dual purposes have an advantage over the con-
ventional ones, The Reflective semiconductor amplifier is an attractive device which
outperforms all the others in its category. The RSOA (invented in 2006 credited to
Samsung Electronics [18]) serves as an optical amplifier, data modulator and also it can
erase the data present in the incoming light. Thus RSOA acts as an excellent candi-
date for bidirectional RoF links. This device eliminates the necessity having an EDFA
14
Figure 2.4: RoF scheme with optical carrier suppression
together with complicated circuitry to erase the data along with the expensive Mach-
Zehnder modulator. All the three actions can be performed in a single device thus it
acts as a compact device highly recommended for incoming light remodulation tech-
nique. This concept was designed in the PON [19] then the same idea was extended to
wireless scheme (RoF).
The figure 2.5 indicates the output power versus input power for a data rewriter.
Here we observe that when the input power exceeds a certain value then by the natural
output characteristic of the SOA the output power appears to be uniform. Thus the in-
coming signal power is crucial in the design of a light remodulation scheme. The slope
of the curve in the saturation region would decide the overall efficiency of the SOA
acting as a light remodulator [20].
15
Figure 2.5: SOA used as a data re-writer
RSOAs offer several advantages [21]such as:
• Low fabrication cost
• Small dimensions
• Easy integration
• Wide bandwidth
• Low insertion loss
• Fast time response
• Colorless
There has been a tremendous interest in schemes involving RSOA. If in the fu-
ture RSOA becomes cheaper than EDFA and other conventional devices, then schemes
designed with RSOA would be a major hit in the market and research alike. Also,
the RSOA operates at a relatively low speed (commercial devices offer 1.25 Gbps)Here
WDM-PON scheme is implemented with necessary changes. Multiple DFB laser sources
are used to generate narrowband light. It is given to a circulator which allows light to
pass from port 1 to 2 only. Similarly, the incoming data is allowed to interact only with
the associated photodiode for capturing the incoming uplink data. All the sources are
given to the AWG, which is one of the most popular devices to achieve WDM (even
16
DWDM) technology.
The multiplexed signal is given to a single SMF, at the destination another AWG is
used which acts as a demultiplexer and eventually segregates the combined light into
various colours. From there it is redistributed using the ONU which in turn has ar-
chitecture similar to the input side of the system. The difference is it has a circulator
connected to a photodiode in one port and the RSOA in the other port. The RSOA
amplifies the incoming light meanwhile it modulates that light (which is erased of the
previously carried data) with uplink data and given to the AWG. Hence both uplink
and downlink are successfully implemented in a very economic way. But this scheme
also has some inherent problems as this cannot be extended to long haul or ultra long
haul applications and also tuning of the RSOA is extremely difficult for customized
networks. In spite of all those problems this scheme has promising scope for future
multigigabit communication systems [19].
Figure 2.6: Bidirectional WDM-PON technology using RSOA
A slight variation of this scheme was used for transmission of broadcasting in the
downlink and normal data communication in the upstream channel. This is a poten-
tial application for analog CATV applications which also require some interactive data
while browsing through the channels. Hence, a low speed upstream data would suffice
in those situations. Again usage of RSOA paves way for good applications but the in-
dustrial applicability and the economic viability of the scheme is yet to be determined.
But this scheme actually motivates the author to explore the applications of RSOA [22].
17
After the successful demonstration of utilizing RSOA in bidirectional links (2006)
lot of other variations have been published. Some of the attractive schemes are dis-
cussed below:
A multiplexer and demultiplexer are used at the transmitter side. The DFB lasers’
outputs are coupled onto Electro-absorption modulators and multiplexed onto a AWG
which has a power splitter, wherein the other half of the incoming power goes to the
demultiplexer at the transmitter side, which sorts out different colours of light and ex-
tracts useful information associated with that particular color. But on the receiver side
only a single demultiplexer is sufficient, because a circulator along with pin photodiode
(in one port) and RSOA (in other port, modulates the uplink data, also provides optical
gain) is sufficient [23]. Though, not a significant paper but useful for exploration of
RSOA in a WDM system.
Proposal of a scheme which caters to the need of CATV users, broadband users,
PON users and wireless users all put together would be the “Holy Grail ”of optical fiber
communication achieved till date. It is a very well known fact that the optical fiber
channel is under-used. Any significant development of technology takes place only
when there is a successful market which consumes the data. But rightnow, there is a
saturation of the market for cellular phone users. Hence, the capitalists are marketing
data consumption like never before.
Nevertheless, achievement of heavy data rate in the optical system is not really un-
realistic because, the optical fiber has extremely large bandwidth ( in the wavelength
zone of communication interest ) out of which only 2-3% is utilized the rest of the
bandwidth is absolutely useless. Thus, those unused portions of the spectrum could
be actually dedicated to the different technological standards. This could be a boon to
customers in residences as well as commercial buildings. One such scheme is the cul-
mination of both wired and wireless technologies [24]. The idea is fairly simple: one of
the CW laser source is modulated by baseband data and the other by millimeter wave
signal. The specific modulation (format and scheme: direct or otherwise) depends on
the budget invested and speed of operation. Both signals are given to a power combiner
and launched over an optical fiber. At the receiver the two colours are segregated and
18
processed. This could be in the bidirectional in nature with the concepts suggested ear-
lier.
Another scheme: simply hires the ideas from other schemes to achieve higher band-
width utilization. Here Fabry Perot is assumed to produce multiple wavelengths. It
should be further noted that reducing the cost of the source is not necessary because the
optical transmitters are naturally cheap. Also, Mode partition noise is a critical element
of interest. However there are several innovative proposals in the scheme worth appre-
ciation.
The incoming different colours are deprived of their carriers and converted to dou-
ble sidebands. The incoming optically suppressed multiple colors are segregated by
the AWG and given to an RSOA which modulates the single colored light with unique
data of that channel and puts a copy onto the sideband of the light (bandwidth wasted)
[7] also it has to be observed that the light signal is amplified, hence works for a fairly
decent distance. This is given to the AWG again to combine the colored light appended
with data. This energy is launched onto the fiber and given to an AWG, acting as a
demultiplexer. At the receiver a conventional scheme is employed. The only attractive
feature is that this scheme culminates wired and wireless technologies. Also the pa-
per is incomplete as the title claims simultaneous bidirectional wired and wireless for a
WDM channel. But experimentally only a single channel with time multiplexed uplink
and downlink data transmission is demonstrated.
Also the necessity of multiple carrier suppression is not thoroughly justified. The
advantage of using a Fabry Perot diode, apart from the obvious monetary benefits, is
not at all stated. Also the experimental demonstration is quite misleading, does not pre-
cisely talk about wired/wireless/simultaneous transmissions. Also the use of two AWGs
could have been easily avoided. No proper justifications to any of the above mentioned
flaws have been reported in this literature [25].
19
Figure 2.7: Simultaneous wired and wireless at 1.25 Gbps
Yet another scheme: with striking simplicity but with worthwhile consequences.
Here the downlink signal is phase modulated and the uplink signal is intensity modu-
lated. Here it is observed that it provides a reasonably good BER and data rate. But the
detection process for the phase modulated signals is complex and uneconomical. This
indicates that researchers need not think only in the device level but can come up with
modulation formats for reliable and sturdy communication link [26].
Cross remodulation is also explored which indicates coupling of a pair of optical
fibers intelligently in order to avoid any distortive effects. Here two AWGs are used for
downlink data itself, one for all wireline signals and the other for all wireless signals.
But this scheme turns out to be highly uneconomical because of four AWGs instead
of two, extra couplings and splices hence reliability of the link is questionable. Also,
the practical deployment of this scheme is messy and not future-proof. But the cross
remodulation concept seems to be a promising solution to a different problem [27].
Culmination of WDM, RoF and PON technologies: Here individual laser sources
are multiplexed using a conventional multiplexer. This combined signal is given to a
MZ Modulator suitably biased which acts as an optical carrier suppressor. Hence all the
incoming colours would lose their carrier signal and two wings appear in the spectrum
of each color.
20
Figure 2.8: cross remodulation
This is again demultiplexed and partitioned into two segments, in other words two
AWGs. The first AWG act on wireless data whereas the second AWG act on the wire-
line data, again an MZM is used to modulate the signal with the relevant data sources.
These signals are in turn multiplexed again. Finally this ensemble of signals is trans-
mitted using an optical fiber, in the base station in each channel a power splitter is used
and one segment: a photodiode is used to extract the information whereas in the other
segment a tunable optical filter is used to understand the information. A similar proce-
dure is carried out in the base station but in the opposite direction.
Here, it is observed that the culmination of the PON and RoF technology is success-
ful. That gives a clear indication that the bidirectional hybrid technologies involving
other well established forms can also be included in the scheme [28].
Sideband routing: Various DSB-SC sources are multiplexed and given to a circu-
lator, from which it is given to a demultiplexer which segregates the light, meanwhile
each color is embedded with the data (both wired and wireless) and this signal is again
appended with the original carrier signal. A similar procedure is used in the base station
to carry out the similar function .
High data rate transmission by frequency quadrupling: The simulation is carried out
for a single channel only. The CW laser is given to a suitably biased MZM at a carrier
frequency of 10 GHz, further more an optical circulator is used and Fiber Bragg grating
21
is used to reflect the carrier back to origin and that carrier signal is actually deviated by
the other port of the circulator. Both of the signals are combined after suitable optical
gain compensation.
This signal is launched onto the optical fiber and given to an MZI one portion of
the signal is used to recover the downlink data from the incoming light signal. The
other part of the light of the MZI is modulated by the uplink signal and launched to the
uplink optical fiber. The Central Office contains a tunable optical filter to extract the
information [29].
Figure 2.9: High data rate Full Duplex RoF scheme
If flexible design is the major concern, then the architecture proposed in [30] is most
suited for high speed links. Here, the optical carriers from various sources are coupled
and launched onto a MZM, which acts not as a conventional modulator but as an optical
carrier suppressor, hence the all the coupled carriers are suppressed and optical millime-
ter waves are generated due to the RF carrier given to the MZM. The flexibility of the
design comes after the optical carriers are suppressed. An optical interleaver is utilized
to segregate the signals as wireless and wireline data, wherein individual MZMs are
used in each channel to put on the baseband data. It must be observed that the baseband
data is embedded onto the optical milli-meter waves.
All the signals carrying data is coupled again and transmitted through the downlink
optical fiber from the central office. An interleaver is again used at the base station to
categorize the ensemble of signals as wired and wireless and those signals are further
22
processed using tunable optical filters and other optical combiners. It must be observed
that the optical carrier reuse is not efficiently carried out in the current scheme due to the
uneconomical use of optical filters and extra MZMs. The proposed architecture would
be suitable for networks which have approximately equal number of wired and wireless
terminals, such as in a university campus.
Figure 2.10: Flexible Full Duplex RoF architecture
The above discussed schemes were adhoc in nature. Now, we consider a specific
link design scheme as reported in [31]. Here a dual port MZM is utilized to provide
a reliable economical RoF link suitable for short haul links. The output carrier from
the CW laser source is given to the LiNbO3 dual port modulator. The two ports are
modulated with different set of signals which in turn is amplified and launched into the
optical fiber. The two ports of the MZM are modulated by a baseband signal of 10 Gbps
and the other port is modulated by a smaller data rate of 2.5 Gbps which is carried by
a 20 GHz RF carrier. Thus the base station (BS) must necessarily have an interleaver
and suitable multiplexer components to successfully extract the information hidden in
the mixed signal.
Though this scheme seems to be exceptional but depends again on the operational
cost in the long run. Also, there is an increased complexity in the detection process
23
at the base station which again adds to the operational cost. But this scheme could be
preferred for high speed data links where speed is a higher priority than the flexibility
or other system design issues. Also the study of this particular link design is not carried
out in the context of a system. Basically this is a generic scheme waiting to be applied
elsewhere.
Figure 2.11: Simulatneous Wired and Wireless RoF link
Schemes in optical generation of millimeter waves
Experimental demonstration of spectrum modification of light using external mod-
ulation is evident due to the inherent nonlinearities in the device. Cascaded configura-
tions are also possible but the problem is bandwidth would not be effectively utilized
[32]. From the various schemes the designer can choose SSB or DSB or other spectral
sharpening schemes.
By cascading the modulators frequency quadrupling is achieved which is extremely
important for generation of millimeter waves again the bandwidth utilization would
be hampered. [33] Frequency octupling is also possible by varying the biases of the
modulators [34].
24
CHAPTER 3
Proposed Scheme
Simulation is the integral part of any proposal. This saves time, money and engineering
man-hours. If and only if the system works reasonably well in the theoretical domain
only then researchers and engineers invest their time in rigging up the actual system.
3.1 Proposed Scheme
The figure below illustrates the general idea of a hybrid network wherein the optical
fiber feeder network bridges the central office and the base station. The terms central
office and base station are used in a general sense. This could mean a data server and
a client or a n ensemble of them which are connected to the either ends of the optical
fiber backbone. The general architecture demonstrates a full duplex operation, the ac-
tual fiber laid out could be a single one (bidirectional fiber, costly) or two fibers (one
for uplink and the other for downlink). The connected nodes would comprise of wired
and wireless terminals.
Figure 3.1: General Block diagram
A more detailed figure is shown below. The Network could mean the humungous
number of terminals connected through the metro network or something likewise. The
downlink and uplink have separate channels for transmission. Each channel is allo-
cated to a particular wavelength chosen for a particular region. The multiplexer then
combines all the colours and launches to the optical fiber backbone. At the receiver the
demultiplexer segregates the colours and extracts useful information. The processing at
the end terminals are usually electronic and are designed on an adhoc basis. The node
would be a combination of wired terminal and wireless terminals.
The wired terminal refers to the computer(s) hooked up to the PON and the wire-
less terminals could mean the cellular phones accessing voice and data services. The
wireless terminals could also be the smartphones and computers accessing broadband
services via the local WiFi. The thick line and the broken line simply illustrates the
difference in electronic processing which is used in the individual terminals.
Figure 3.2: General hybrid architecture
A DFB laser diode of finite linewidth is modulated with a dual drive MZM. One of
the ports is modulated with wireless data, i.e., the baseband data is mixed with a suitable
RF carrier depending on the desired baseband datarate. This frequency translation in
the electrical domain is performed in order to simplify the antenna design.
26
Figure 3.3: RoF link
The other port is modulated with the wired baseband data, originating from the pas-
sive optical network terminal. Both these signals are embedded onto the CW laser light
for a specific channel. The emerging output from the dual port MZM is given to a bidi-
rectional wavelength division multiplexer. A circulator is used at the MZM in order to
avoid optical interferences from the base station side.
The incoming signal from the base station is given to a PIN diode after wavelength
segregation (emanating from the multiplexer). The circulator design is economical for
a full duplex system. The same design is utilized for other channels as well; hence the
WDM central office is realized.
The combined signals of different wavelengths are given to the optical transmission
channel, from which a demultiplexer is used to segregate the wavelengths as designed
in the central office.
Once each optical signal of a particular wavelength is extracted, it is given to a
power splitter which leads to three paths namely PIN diode with mixer, PIN diode and
RSOA.
The PIN diode with mixer extracts the wireless data whereas the PIN diode extracts
the wired data. The remaining portion of the optical signal is given to the RSOA which
27
is operated in the gain saturation region thereby serving the purpose of both amplifier
and modulator; as a result the RSOA remodulates the incoming signal with uplink data.
Figure 3.3 illustrates the idea.
The uplink data is basically baseband data, hence generic in nature; to achieve wired
and wireless signal transmission at the base station side then time division multiplexing
or frequency division multiplexing in electrical domain could be carried out.
3.2 Simulation Setup
The above proposed conceptual scheme was built and tested in OptiSystem 11.
The CW laser with linewidth of 10MHz operating at wavelength of 1552.52 nm, this
was used for channel 1, the subsequent channels were placed at 100 GHz from chan-
nel one’s carrier frequency (193.1 THz). Four channels were chosen for simulation
purpose. Since the proposed architecture is generic in nature the number of channels
could exceed more than eighty, as in DWDM networks. But the availability and precise
linewidth of laser diode array would be difficult to manufacture.
A pseudo-random number generator was used to drive an NRZ pulse generator to
generate the baseband signals; the bit rate would be set accordingly. The PRBS signi-
fies the random source of data such as speech, music, pictures or video. The NRZ pulse
generator was chosen in order to have a simple low speed electrical network. The data
rate was set at 2.5 Gbps for simulation of downlink, and the data rate was set to 1 Gbps
for simulation of uplink. The global parameters were set to match the data rates, thus
we obtain the graphs with reasonably good clarity.
Port 1 of dual drive dual port MZM was driven by an amplitude modulator, with
a carrier frequency of 20 GHz, which in turn was driven by 2.5 Gb/s baseband signal.
The amplitude modulator would act as a mixer, which acts as a frequency translator of
the baseband to the RF clock frequency. Since, the RF carrier could also be extracted
at the base station; simpler electronic circuitry is required apart from the electronic am-
plifier.Port 2 of dual drive dual port MZM was driven by 1 Gb/s baseband signal. The
28
combined signal was given to a bidirectional circulator.
The incoming signal was given to a PIN diode followed by a low pass Bessel filter
to extract the uplink signal, also the uplink receiver subsystem houses the BER ana-
lyzer to plot the eye diagrams . The outgoing signal from the circulator is given to a
bidirectional multiplexer with suitable number of channels, the output of which is given
to a bidirectional SMF of 23km. Since, we focus on short haul networks the reasonable
length of 23km is justified. Also, it was observed that with increased distance the would
be operational but at a lower data rates for both downstream and upstream links.
This was further given to demultiplexer (due to non-availability of a bidirectional
demux an extra mux was used for demonstration purpose). Each optical channel was
given to a power splitter.
One portion is given to a PIN diode followed by a low pass filter; this branch extracts
the wired downstream data. Other portion is given to PIN diode followed by amplitude
demodulator with a carrier frequency of 20 GHz (frequency translator in the electrical
domain) and the remaining power from that optical channel is given to the RSOA, which
remodulates the signal with uplink data (1 Gb/s).
29
Figu
re3.
4:Pr
opos
edsc
hem
e
30
3.3 Results and Discussion
The global parameters of OptiSystem were suitably set to visualize the results.
Fig 3.5 Illustrates the optical spectrum of a single channel after modulation by the
downstream signals (wired and wireless). The extremely high data rate is limited by the
speed of operation of the MZM, thus in practical design there is a trade-off between cost
of installation and the operational data rate. The optical carrier suppression technique
might appear to be economic since the cost is shared between a lot of channels, but
the current scheme offers high bandwidth signal transmission for two different types of
data sets. Hence, the proposed scheme is preferred for a networks which have multiple
terminals (wired and wireless) of approximately equal proportion.
Figure 3.5: CW laser spectrum after modulation
Fig. 3.6 shows the optical spectrum after multiplexing of four channels each sep-
arated by 100 GHz. This separation is capable of supporting huge number of users
and high data rate. The power levels should be carefully adjusted in order to prevent
interchannel distortion and other nonlinearities. If a long haul extension of the link is
considered then transients in the optical amplifier also has to be taken into account, opti-
mal power level should be set depending on the designed distance. For, short haul links
highly sensitive and high bandwidth PIN diodes would serve the purpose of limiting the
launched power.
31
Figure 3.6: WDM signal spectrum after modulation
The eye pattern for downlink wireless data is shown in Fig 3.7 for a data rate of 2.5
Gb/s. The optical fiber length was set to 23km, which is a typical distance for covering
a typical office campus. The input power was iterated from -10 dBm to 0 dBm in steps
of 0.2 dBm. All the above said manipulations were performed on a single link within
the WDM system. The concept of increasing power of the input laser source could be
conceptually corresponded to decrease in the fiber length, due to the natural attenuation
encountered in the fiber. For a desired input power level, data rate, fiber length and
specified BER. The eye opening is decent for a reliable communication link.
Figure 3.7: Eye diagram wireless 2.5Gbps
32
Fig. 3.8 demonstrates the eye pattern for downlink for wired signals operative at
1 Gb/s. The optical fiber length was set to 23km. This suggests that the terminals of
PON are operating at a lower datarate than its wireless counterpart; hence this scheme
would be suitable where high speed mobile terminals have a higher priority than the
PON terminals. The eye pattern seems to validate the proposed scheme to be deployed
in the real world.
Figure 3.8: Eye diagram for wired downlink
Also fig. 3.9 shows log(BER) versus received optical power. This graph proves
that a reasonably good BER could be achieved at a relatively low operative power for a
single channel.
Figure 3.9: log BER versus received power
33
3.4 Applications
Intra campus network which could be used for
• High quality raw Music sharing: but this also requires high quality speakers forgood sound throughput.
• High resolution photo sharing: The cameras could be connected with wirelessmodules to transfer the HQ photos directly to the network.
• Multiple simulatneous video conferencing: Reasonably good quality video con-ferencing could be carried out with the proposed scheme. Also, a limited numberof multiple video conferencing could be supported with a complex electronichardware.
• Multiplayer high end graphic intensive gaming: Graphic intensive gaming re-quires very high data rates to be supported, thus a couple of gamers could beentertained based on the specific design.
• Storage network: Optical storage networks are becoming popular which againrequires high data rates and a mixture of technologies.
• Broadcasting of intra-campus programmes within the campus
• Interactive televison: which again requires a full duplex network with high dataratedownlink but a comparitively lower uplink datarate.
• Amalgamation of CATV with digital services could also be designed.
• Augmented Reality also requires tremendous amounts of bandwidth, hence theproposed scheme could be used for educational purposes.
34
3.5 Conclusion and future wok
A novel architecture for WDM-RoF amalgamated with wired network is proposed.
The scheme is relatively inexpensive and converges multiple technologies with the same
optical backbone network. Also, we observe that it provides a full duplex signal trans-
action. The theoretical investigation shows that the proposed scheme is capable of
offering reliable communication at reasonably high data rates. The simulative analysis
demonstrated a 2.5 Gbps downlink and 1 Gbps uplink, which is suitable for short haul
networks. The proposed scheme requires a high investment for installation but proves
to be highly economical when the number of subscribers increase.
The scheme could be tested on a real campus. Flexible design for the proposed
scheme could be explored which is future-proof compared to the proposed network.
The data-rate could be increased but with a slight detorioration of the eye diagram and
BER parameters.
35
List of Publications
[1] “A Novel Architecture for Full Duplex WDM Radio Over Fiber System Con-
verged With Wired Network Using Dual Drive MZM and Light Remodulation using
RSOA ”accepted for Ninth International Conference on Wireless and Optical Commu-
nications Networks WOCN 2012 conducted by IIT, Indore. September, 2012
[2] “Simulative Analysis of a WDM RoF Architecture for Full Duplex Wired and
Wireless Network ”accepted for National Conference on Technological Trends NCTT
2012 conducted by College of Engineering, Trivandrum. August, 2012
36
37
38
REFERENCES
[1] Jeff Hecht, 1997 City of Light: The story of Fiber Optics, Oxford University Press
[2] Kao, K. C., Hockham, G. A. ,“PDielectric-fibre surface waveguides for optical fre-
quencies”Proc. IEE 113,1966 1151-1158.
[3] Rajiv Ramaswami “Optical fiber communication: from transmission to network-
ing”, IEEE communications magazine 2002.
[4] Wireless communication and networks, presentation by Aglar, Bahar 2007.
[5] http://en.wikipedia.org/wiki/List of countries by number of mobile phones in use
[6] Seeds, A. J.; Williams, K. J.; , “Microwave Photonics,”Lightwave Technology,
Journal of , vol.24, no.12, pp.4628-4641, Dec. 2006, 96 citations.
[7] Harvey, C.; Symington, I.C.; Kirsten, D.; , ‘communications utilising radio over
fibre techniques for the local loop,”Communications, 1991. ICC ’91, Conference
Record. IEEE International Conference on , vol., no., pp.1171-1175 vol.3, 23-26
Jun 1991.
[8] Govind P Agrawal, 3rd ed, 2002, Fiber-Optic Communication Systems, John-Wiley