Proposal for Towards Freeband Communications –
Research Program on Telecommunication
Administrative data
Title of the proposal:
High capacity multi-service
in-house networks using mode group diversity multiplexing
Proposal for Towards
Freeband Communications – Research Program on Telecommunication
Proposers:
Prof.ir. A.M.J. Koonen (project leader),
COBRA Institute
Group Electro-Optical Communication Systems (ECO)
Eindhoven
University of Technology
Den Dolech
2
5612
AZ Eindhoven
tel. 040 2474806 (secr. 3451)
fax 040
2455197
e-mail a.m.j.koonen@tue.nl
pProf.dr.ir.
J.W.M. Bergmans,
Group Signal Processing Systems (SPS)
Eindhoven University of Technology
tel. 040
2474438
fax 040
2433066
e-mail j.w.m.bergmans@tue.nl
iIr. H.P.A. van
den Boom,
COBRA Institute
Group Electro-Optical Communication Systems (ECO)
Eindhoven University of Technology
tel. 040
2473444
fax 040
2455197
e-mail h.p.a.v.d.boom@tue.nl
dDr.ir. F.M.J. Willems,
Group Signal Processing Systems (SPS)
Eindhoven University of
Technology
tel. 040
2473539
fax 040
2433066
e-mail f.m.j.willems@tue.nl
dDr.ir. M.J. Bastiaans,
Group Signal Processing Systems (SPS)
Eindhoven University of Technology
tel. 040
2473319
fax 040
2433066
e-mail m.j.bastiaans@tue.nl
Ddr.ir. P.C.W. Sommen.
Group Signal Processing
Systems (SPS)
Eindhoven University of
Technology
tel. 040 2473634
fax 040 2433066
e-mail p.c.w.sommen@tue.nl
Context (Inpassing)
The proposed project fits in the scope of BraBant BreedBand (B4), the joint industry-university research alliance on broadband network techniques initiated by Eindhoven University of Technology, KPN and Lucent Technologies. It also anticipates on the proposed new IOP framework programme “Generieke Communicatie” addressing communication in access and residential networks.
Applications elsewhere (Aanvragen
elders)
None.
Keywords (Trefwoorden)
Multimode fibre, Polymer Optical Fibre, in-house networks, electrical signal processing
Domain and Research topics (Domein en
onderzoekstopics)
Freeband area
2: Domain of the first mile (between user and network),
addressing
research topics on compatible networks in a residential/office environment,
development of new infrastructures, coding and modulation methods, system
design, physical structure of the network in local and residential/office
areas, transparency and seamless usage, handling of multi-modal data streams on
and in home/office network environments
Strengthening of
infrastructure
(Versterking infrastructuur)
The
proposed project will strengthen the expertise in the COBRA institute and in
the B4 alliance regarding residential networking. It will promote the
development of broadband campus infrastructures at the Eindhoven University of
Technology, and thus support the university’s goal to realise a widespread
digital learning environment.
1 Project summary
1.1 Research
Now, two PhD. positions are available for this project.
Multimode fibre, and in
particular Polymer Optical Fibre (POF), has emerged as a very attractive medium
for realising transparent easy-to-install in-house communication networks. It offers distinct
advantages in comparison to traditional fixed wiring media such as coaxial
copper cable and twisted pair copper cable: it offers significantly more
bandwidth and lower losses, and more importantly complete transparency to
signal format and protocol. However, in comparison to standard single-mode
silica fibre, POF has per unit of length a higher attenuation, and a lower
bandwidth (due to its higher dispersion caused by its multimode waveguiding
behaviour). Therefore its applicability for broadband data communication is
limited.
In this project, mode group diversity
multiplexing is proposed as an approach to increase the
fibre’s data transport capability. With selective mode group launching, subsets of the
large volume of modes are being deployed individually for data transport, and at the
receiver side adaptive
electrical
signal processing is used for removing the crosstalk (caused by mode mixing
in the fibre) between these channels. Thus several independent
communication channels can be created, and multiple services can be integrated independently of each other in this single fibre infrastructure. Its
functionality is comparable to wavelength multiplexing, but without requiring
the more costly wavelength-specific sources and wavelength (de-)multiplexing
system functions. Mode group diversity multiplexing may outperform wavelength
multiplexing, when the costs of the electrical signal processing required are
lower than the costs for wavelength multiplexing.
The project will analyse and characterise in-depth the behaviour
of multimode (polymer) optical fibre under selective mode group excitation,; design and realise mode-group selective
transmitter and receiver system modules; analyse, design and realise optimum electrical
signal processing circuits for inverting the mode mixing processes and
safeguard adequate transmission performance; and integrate these system modules
in a laboratory testbed for verification and evaluation of the system’s
transmission performance. Extensions into bi-directional system operation
and point-to-multipoint system operation will be studied as well.
1.2 Utilisation
The main goal of the
project is to architect, design and validate a universal future-proof in-house
network for seamless transparent communication within the residential area,
between the various in-house user terminals. Currently available technology for in-house
communication is expensive, complex to install and user-unfriendly, and lacks
forward compatibility and standardisation. Thus ICT innovations in the consumer
market are slowed down. Seen the fast growing demand for widely varying
communication services in the user domain, on short notice solutions need to be explored for a universal transparent
broadband in-house infrastructure capable of hosting a wide variety of services
with different quality-of-service demands. Multimode (polymer) optical fibre may provide such a universal broadband
infrastructure,
but should then provide cost-effective multiplexing techniques to enable
simultaneous use of multiple services. Mode group diversity multiplexing can be a promising solution
for this. In combination with a residential gateway (as studied in complementary
other projects executed by the proposers at TU/e, and being defined in
the upcoming IOP framework programme “Generieke Communicatie”), this universal
in-house network can flexibly interact with the various public access networks.
Various industrial
partners have expressed interest in the results foreseen from this project, and have indicated
promising prospects in their market segments. Philips is interested in the signal processing
techniques
(related to their PDIC products), Draka Fibre in the extended application
possibilities of multimode fibre (in particular their silica high-performance
fibre), and
KPN, Lucent and Philips in the novel in-house networking possibilities and interaction with residential
gateways and public access networks.
1.1Samenvatting
Multimode vezels en in het
bijzonder polymeer optische vezels zijn uitgegroeid tot een zeer aantrekkelijk
medium om transparante gemakkelijk
te installeren in-huis
communicatienetwerken te realiseren en die
gemakkelijk zijn te installeren. Deze vezels bieden
duidelijke voordelen vergeleken met traditionele media voor vaste bedrading
zoals coaxiale koperkabel en twisted pair koperkabel: de vezels het leverent significant meer bandbreedte en lagere verliezen,
en nog belangrijker de vezels zijnhet is volledig transparant voor signaal format en protocol. Echter, vergeleken met
standaard single-mode glasvezel heeft polymeer optische vezel een hogere demping
per lengte eenheid en een lagere bandbreedte (door dezijn grotere dispersie die wordt veroorzaakt door hetzijn multimode gedrag). Daardoor is de toepasbaarheid
voor breedbandige datacommunicatie beperkt.
In dit project, wordt mode
group diversity multiplexing voorgesteld als een oplossing om de
datatransportcapaciteit te vergroten. Met selectieve mode groep
inkoppeling worden subsets van het
grote aantal modi individueel gebruikt voor datatransport, en bij de ontvangkant
wordt adaptieve elektrische signaal processing gebruikt om de overspraak tussen de
kanalen (veroorzaakt door mode mixing in de vezel) te verwijderen. Er kunnen
dus verschillende onafhankelijke communicatiekanalen worden gecreëerd, en
meervoudige services kunnen onafhankelijk van elkaar worden geïntegreerd in
deze infrastructuur met één vezel. De functionaliteit is vergelijkbaar met
golflengte multiplexing. Maar dan zonder de duurdere golflengtespecifieke
bronnen en golflengte (de-)multiplexing systeemfuncties.
In het project zalal het gedrag van multimode (polymeer) optische
vezels uitgebreid worden geanalyseerd en gekarakteriseerd met selectieve mode
groep aanstraling; zullen een mode groep selectieve zender en ontvanger modules
worden ontworpen en gerealiseerd; zullen optimale elektrische
signaalprocessing circuits worden geanalyseerd,
ontworpen en gerealiseerd om de mode-mixing processen te inverteren en
adequate transmissieperformance te bewaken worden
geanalyseerd, ontworpen en gerealiseerd; zullen deze systeemmodules worden
geïntegreerd in een laboratorium testbed voor verificatie en evaluatie van de
performance van de transmissie van het systeem. Uitbreidingen voor
bi-directionele transmissie en punt naar multipunt transmissie zal ook worden
bestudeerd.
1.1Utilisatie
Het hoofddoel van het
project is het bouwen, ontwerpen en valideren van een universeel toekomstvast
in-huis netwerk voor transparante communicatie tussen de verschillende
terminals van gebruikers in-huis. Technologie die momenteel beschikbaar is voor
in-huis communicatie is duur, moeilijk te installeren en
gebruikersonvriendelijk en het ontbreekt aan compatibiliteit en
standaardisatie. ICT innovatie gaat dus
langzaam in de markt van de consument. Gezien de snelgroeiende vraag naar verschillende
communicatieservices in de omgeving van de gebruiker moeten op korte termijn
oplossingen worden onderzocht voor een universeel transparante breedbandige
in-huis infrastructuur, die in staat is om een grote variëteit van services met
verschillende kwaliteitseisen te ondersteunen. Multimode (polymeer) optische
vezel zou een dergelijke universele breedbandige infrastructuur kunnen bieden,
maar er zou dan wel een kostenefficiënte multiplextechniek moeten zijn om
gelijktijdig gebruik van verschillende services mogelijk te maken. Mode group
diversity multiplexing zou een
veelbelovende oplossing kunnen zijn. In combinatie met een residential gateway
(zoals wordt bestudeerd in complementaire andere projecten die worden
uitgevoerd bij de TU/e door de indieners bij de TU/e, en worden gedefinieerd in het nieuwe IOP
programma “Generieke communicatie”), kan dit universele in-huis netwerk
flexibel samenwerken met de verschillende toegangsnetwerken.
Diverse industriële partners
hebben hun interesse uitgesproken voornaar de resultaten van dit project en hebben
veelbelovende vooruitzichten uitgesproken voor hun marktsegment. Philips is
geïnteresseerd in de signaal processing technieken (in relatie met hun PDIC
producten), Draka Fibre in de uitbreiding van de toepassingsmogelijkheden van
multimode vezels (in het bijzonder hun high-performance glasvezel), en
KPN, Lucent en Philips in de nieuwe
in-huis netwerkmogelijkheden en de interactie met residential gateways en
publieke toegangsnetwerken.
1
Research team
1.1
Staff involved in the
project
Name Position Time Field
Ir. H.P.A. van den Boom Assist. Prof. 0.2 fte Optical
fibre system technologies; daily guidance
Ing. F.M. Huijskens Technician 0.3 fte Optical
and electrical hardware construction and test
Prof.ir. A.M.J. Koonen Full prof. 0.1 fte Broadband access systems; group leader and
promotor
Dr.ir. F.M.J. Willems Assoc. prof. 0.1 fte Signal coding; daily guidance
Dr.ir. M.J. Bastiaans Assoc. prof. 0.1 fte Signal analysis
Dr.ir. P.C.W. Sommen Assoc. prof. 0.1 fte Adaptive signal processing
Prof.dr.ir. J.W.M. Bergmans Full prof. 0.1 fte Signal processing; group leader and promotor
Applied
for at Freeband PhD student 1 fte Mode
group diversity multiplexing - optical system analysis, design, realisation and
system experimentation
Applied for at Freeband PhD student 1 fte Mode group diversity multiplexing – signal
processing analysis, deisgndesign,
realisation and system experimentation
1.1
Candidates
No candidates for the PhD positions have been
selected yet; however, many applications of well-qualified students regularly
come in.
2 Scientific description of the project
2.1 Research program
2.1.1 Introduction, problem statement
Optical fibre is an excellent medium for providing high bandwidth and service format transparency. It has brought tremendous data transport capabilities in core telecommunication networks, and it is now conquering the metropolitan area networks and subsequently it is moving into access networks. Coming closer to the end user and penetrating his residential area, however, the costs of installing and maintaining the fibre network become ever more important. Polymer Optical Fibre (POF) offers distinct advantages regarding cost-effective installation and maintenance in in-building business and residential areas, as compared with the commonly used single-mode fibre. The POF’s large core considerably eases coupling and splicing, and its flexibility and ductility enables fast installation in often less accessible customer locations. It offers also distinct advantages in comparison to traditional fixed wiring media such as coaxial copper cable and twisted pair copper cable: it offers significantly more bandwidth and lower losses, and more importantly complete transparency to signal format and protocol. Similar advantages are obtained with silica large-core fibres, although these fibres are less flexible and ductile and thus less easy to install.
However, in comparison to standard single-mode silica fibre, POF has per unit of length a higher attenuation, and a lower bandwidth (due to its higher dispersion caused by its multimode waveguiding behaviour). Therefore its applicability for broadband data communication is limited.
POF material compositions and index profiles are being explored to improve the fibre characteristics. Perfluorinated (PF) polymers have reduced losses to below 20 dB/km at 1300 nm wavelength, 30 dB/km at 800 nm, and 60 dB/km at 650 nm. Graded-index PF fibres (PF GI-POF) have surpassed bandwidth×length products of 1 GHz×km. In conjunction with these improvements in polymer fibre technology, research is needed regarding system techniques to extend the data transport capabilities of POF-based networks.
A lot of progress has been made already both on the POF characteristics and on high-speed POF system techniques [[1]] [[2]]. Recently, the TUE-ECO group has reported a Gigabit Ethernet PF GI-POF link with a length of about 1 km [[3]]. However, to further advance the capabilities of POF-based networks, new signal transport techniques have to be devised.
In this project, mode group diversity multiplexing is proposed as an approach to increase the fibre’s data transport capability. It also enables to create several independent communication channels in a single fibre infrastructure, and thus allows to integrate multiple services independent of each other in this infrastructure. Its functionality is comparable to wavelength multiplexing, but without requiring the more costly wavelength-specific sources and wavelength (de-)multiplexing system functions. Mode group diversity multiplexing may outperform wavelength multiplexing, when the costs of the electrical signal processing required are lower than the costs for wavelength multiplexing.
2.1.2 Proposed research
To exploit the
ease of installation of POF and at the same time to enable high-capacity data
transport, it is proposed to investigate a novel system technique to extend the
system capabilities of POF-based networks: mode
group diversity multiplexing, in which the many modes guided in a POF are
subdivided in groups, and each group is employed as a transmission channel. Thus
several independent channels are created in a single fibre, offering the same
functionality as can be achieved with wavelength division multiplexing (WDM).
Mode group diversity multiplexing may, however, outperform wavelength
multiplexing if the electronic signal processing required can be implemented
cheaper than the WDM circuitry. Moreover, sets of mode-group-diversity
multiplexed signals may be stacked in a next hierarchical level with wavelength
multiplexing, thus creating independent service groups in a unified POF-based
transparent in-house network. Thus e.g. broadband user terminals with
fixed-wire connections (such as Gigabit Ethernet) may be supported, as well as
broadband wireless terminals (such as in a wireless LAN); this method of
service integration is illustrated in Figure 1Figure 1Figure 1Figure 1Figure 1Figure
1.
The proposed research activities are grouped in two closely related fields: those addressing optical system aspects, and those addressing electrical signal processing aspects.
Figure 11111 Transparent In-House
Network with Polymer Optical Fibre, integrating fixed-wired and wireless
services in a single infrastructure
2.1.2.1 Mode group diversity multiplexing – optical system aspects
From an information theory viewpoint,
multimode fibre has a larger information transport capacity than single-mode
fibre, as it guides a multitude of modes. However, usually many modes are
excited simultaneously, and due to mode dispersion this results in a much lower
transport capacity. Selective mode group launching may considerably increase
the bandwidth [[4]], [[5]].
An up to fourfold increase of the bandwidth has been reported by injecting a
small light spot radially offset from the fibre core centre, exciting less than
50% of the fibre modes [[6]]. Provided only a
limited amount of mode mixing occurs, as is the case in high-quality multimode
fibres, one may increase the fibre transport capacity by mode group diversity
multiplexing [[7]]. As shown in Figure 3Figure 2Figure 2Figure 2Figure 3Figure
2, N mode groups are selectively excited in parallel with separate
data streams from N laser diodes. At
the receiving end, M ³ N detectors produce M signals which each contain a mix of
the N input data streams due to the
mode mixing incurred in the fibre. A signal processing circuit subsequently
inverts the N x M
transmission matrix (using a training sequence of known bits). Thus the M received streams are decorrelated, and
the N original data streams are
recovered. Experiments have shown the technical feasibility at 1 Mbit/s
per channel for N=M=2; but the technique should be
scalable to Gbit/s rates with appropriate high-speed processing circuits.
Figure 322232 Mode group diversity
multiplexing
It is proposed to scale up the technique by deploying higher bitrates and higher diversity factors N and M using selective mode group launching. Similarities with space-time coding such as exploited in wireless LANs will be investigated. Firstly, experimental characterisation of graded-index POF will be done regarding mode mixing and the impact of selective mode group launching on the fibre’s bandwidth. A model for describing the fibre’s behaviour under limited launching will be set up, taking the characterisation results into account. It will be implemented in a system simulation tool. Subsequently, a system design will be made encompassing a multi-laser transmitter end including driver circuitry, and a multi-photodiode receiver end. Requirements for the high-speed signal processing circuitry will be defined, in order to obtain a sizeable increase in fibre transport capacity. The system’s performance will be analysed by means of simulations. A small-scale laboratory point-to-point optical system setup will be built to verify the technical feasibility.
Furthermore,
the mode group diversity multiplexing may actually, after signal processing,
enable a number of transparent communication channels in parallel on the same
POF, thus allowing to host services with different signal formats into a single
infrastructure. An analysis will be made how sensitive the system is for pertubationsperturbations
in the optical system parameters; given an optimum setting of the transfer
matrix parameters in the signal processing circuit, system simulations will be
run to assess how the system performance is affected when the system parameters
such as laser linewidth are varied.
The mode group diversity multiplexing approach offers basically the same functionality as wavelength multiplexing and subcarrier multiplexing, and techno-economic studies will be done to assess the pros and cons of mode group diversity multiplexing versus these two alternative techniques.
Next, it will
be studied how the concept can be extended to a bi-directional system. The
inverted transmission matrix data deployed by the signal processing circuit and
obtained with the downstream training sequence procedure,procedure may
facilitate the selective launching for upstream transmission.
The mode group diversity approach can also enable new medium access control strategies in point-to-multipoint architectures. The potential for this will be analysed.
Channel Characterization
a) experimental approach
A
computer-controlled set-up will be realized for the measurement of the
intensity distribution over the surface at the receiving end of the multimode fiberfibre, using
different launching conditions at the transmitting end. It has been reported
recently how this near field intensity pattern in a POF depends on the
launching conditions, thus basically showing how different mode groups can
constitute different communication channels; see Figure 5Figure 3Figure 3Figure 3Figure 5Figure
3 [8]. The influence of the following parameters
will be measured: place of the launching beam at the end surface of the fiberfibre, spot size
of the launching beam, angle of the launching beam with respect to the end
surface of the fiberfibre, length of the fiberfibre, step index
and graded-index fiberfibre, silica and polymer optical fiberfibre. Moreover,
the influence of bending and stress on the fiberfibre will be
measured. The results of these measurements can give a first approach of an
optimum launching and an optimum detection array. By means of a separate
set-up, bandwidth measurements of the different channels using different mode
groups will be carried out.
Figure 533353 Near-Field Patterns at output of multimode POF, depending on
launching conditions [[8]]
a) Exciting all modes, resulting in NFP spread out over whole fibre core
b)
Exciting low order modes only; NFP
clearly focussedfocused in
core centre
c) Exciting high order modes only; NFP concentrated in ring shape close to core-cladding boundary
b) theoretical approach
Since a multimode fiberfibre behaves like
a geometric-optical system, there is a great advantage in representing the
optical field propagating through such a fiberfibre by its
Wigner distribution function. The real-valued Wigner distribution function
represents an optical field - like in radiometry - in a phase space, i.e. in a
combined space/spatial-frequency domain. The propagation of partially coherent
light through a fiberfibre is then governed by a partial
differential equation, and the propagation has a conceptually clear
interpretation: the amplitude of the Wigner distribution function along the
characteristics of the partial differential equation (i.e., along the
geometric-optical ray paths) remains invariant. In particular for rotationally
symmetric fiberfibres, the partial differential equation has
a reduced number of variables, thanks to the two ray invariants. (see [[9]],
in particular Sections 5B and 8E; or [[10]],
in particular Sections 6 and 8). We would like to investigate the possibility
of using the Wigner distribution function to determine the propagation
characteristics of the multimode polymer and silica fiberfibre.
Another phase-space description -
the Gabor expansion - of the optical field is worthwhile studying as well.
Gabor's signal expansion decomposes the optical field in a set of shifted and
modulated versions of a so-called elementary signal. As such an elementary
signal, one could think of a Gaussian function (where we can benefit from the
large knowledge about the propagation of Gaussian beams through linear optical
systems) or think of the most significant mode that propagates through the fiberfibre. Once the
elementary signal (or synthesis window) has been chosen, an analysis window can
be found with the help of which Gabor's expansion coefficients (see [[11]],
in particular Section 14.3).
2.1.2.2 Mode group diversity multiplexing – electrical signal processing aspects
Analysis of system capacity and signal processing strategy
Modeling the physical behaviour of the
optical fiberfibre should provide the answer to several
questions.
Figure 744474 ModellingModeling the optical
system transfer function (for N=M=2)
Given the N transmitters and M detectors, there is an NxM matrix of which element c(n,m) describes the signal transfer from
transmitter n to detector m ; Figure 7Figure 4Figure 4Figure 4Figure 7Figure
4 illustrates this model for
the case N=M=2. These complex elements c(n,m)
have both an amplitude (representing attenuation) and a phase (representing
phase delay) of the specific path (n,m). Due to random fluctuations in the
mode coupling, fluctuations in the complex elements c(n,m) will occur. When the optical sources have a coherence time
larger than the differential delay times between the modes, also speckle
patterns caused by interference between the modes will contribute to
fluctuations in the c(n,m)
values [[12]].
Methods to deliberately reduce the coherence time of the sources to reduce the
contrast of these speckle patterns need to be devised.
It is important
to know the statistical properties of the transfer matrix elements and the
statistics of the noise that is picked up by the detectors. Based on these
statistical properties we can first determine the capacity of the medium, or at
least find good bounds for it. Knowledge of the capacity gives us guidelines in
designing coding techniques that attempt to approach this capacity. Crucial is
of course that the coding complexity remains acceptable. As an example we
mention the sequence of developments in multiple-transmit and multiple-receive
antenna systems in a Raleigh fading environment. The capacity of such a system
was determined by Teletar [[13]]
and Foshini and Gans [[14]]. This capacity increases dramatically
with the (minimum of the) number of transmit and receive antennas. Tarokh et al. [[15]]
proposed subsequently so-called space-time codes for
these channels. Afterwards space-time codes and signal processing techniques
have been suggested for wireless local loop applications (Lucent’s BLAST
project [[16]] ). A similar sequence of
research steps is needed for our problem setup, transmission over a multi-mode
optical fiberfibre.
So far we
discussed here only behavior of the multimode fiberfibre in the
spatial domain. We have ignored temporal dispersion and jitter. However since
we need high data rates we cannot hope that these impairments are
insignificant. Knowing the statistical properties of the dispersion for all
modes and the jitter is essential. These properties will have their effect on
the capacity of the system and will determine the coding techniques that are
required. It should be mentioned that in general a trade-off can be made
between synchronization and equalization on one hand and coding on the other
hand.
A third
important issue is that of feedback (see Figure 3Figure 2Figure 2Figure 2Figure 3Figure
2). It is of greatest
importance whether or not the transmitter knows the parameters of the actual
channel. In that case the code (signaling technique) can adapt to this
knowledge and e.g. privilege the good channels with respect to the bad
channels. When this knowledge is not available redundancy has to be added to
the signals. With this redundancy errors (or erasures) can be corrected. It
will be clear that in general the capacity of a system is larger if the
transmitter knows these channel parameters. Feedback is necessary to inform the
transmitter about the channel. Note that it is not so hard for the receiver to
determine the actual channel parameters; it can send the parameters back to the
transmitter by a relatively low-speed return channel. Concepts for this
feedback channel will be studied in depth, and realised and validated in an
experimental setup.
It is our objective to characterize the multi-mode channel behavior in the first year of the project. In the second year capacity calculations will be performed and the signaling method will be selected. In year three attention should be given to the signal processing techniques at the receiver, i.e. synchronization, equalization, and error correction. System integration will take place in the last year of the project.
Electrical signal processing
system
In order to perform the electrical signal processing steps required for
the mode group diversity demultiplexing process, a system as exemplified in Figure 8Figure 5Figure 5Figure 5Figure 8Figure
5 is required. This
system consists of a photo detector IC (PDIC) followed by a signal processor
equipped with an analog-to-digital (A/D) converter card and several FPGA cards.
The multi-channel analog-to-digital converter samples the parallel output
stream of the PDIC, i.e. one analog signal for each detector. Then a first
high-density field programmable gate array (FPGA) takes care of the
synchronisation and (adaptive) array equalisation, performing the transmission
matrix inversion. A second FPGA performs the error detection and correction,
and produces estimates of the transmitted symbols. Because of the signalling
rates required, the speed of the cards and the power of the overall processor
should be as high as possible. In the first phases of the research
synchronisation, equalization, and error detection plus correction will be
carried out in software. This requires a high-speed disk to store the output produced
by the A/D converter. This data can then be processed off-line afterwards. In
later phases, when it is clear which signalling methods will be applied, the
corresponding decoding methods and the synchronization and equalisation
techniques will be implemented in FPGA-s. These experiments must eventually
lead to a low-cost design (ASIC).
Note that the
equalizer can be trained in the initial transmission phase by means of special
signal sequences sent by the transmitter. Also, when the error detector signals
an error rate exceeding a certain level, via the feedback channel a new
training period can be started at the transmitter end to re-initialise the
system. Moreover, during the transmission process, the equalizer is adapted
regularly by the control information that is produced by the error-correcting
decoder.
Figure
855585 Signal processing at the receiver site (example)
The design of
the transmitter is essentially simpler than that of the receiver; Figure 10Figure 6Figure 6Figure 6Figure 10Figure
6 gives an example of a
possible setup. In the non-feedback case, the data is encoded, however only
procedures are permitted then that allow decoding at a very high data rates.
Therefore encoding cannot be very complex. In the feedback case the transmitter
will be slightly more complex since the signaling procedure may depend on the
state of the medium, but again the receiver will perform the major part of the
processing. Via the feedback channel, also a new training period to re-initiate
the system can be started, as mentioned above.
Figure 10666106 Signal processing at
the transmitter site (example)
2.1.2.3 Mode group diversity multiplexing - System integration
After realising
the optical system modules and the electrical signal processing system, and
having tested these separately, the modules will be integrated in a
uni-directional mode group diversity multiplexing system (with a return
feedback channel for initialisation and control), according to the basic setup
shown in Figure 3Figure 2Figure 2Figure 2Figure 3Figure
2. The system will undergo
end-to-end testing, and its performance parameters will be evaluated (power
budget, bit error rate, capacity, robustness against fibre movements and changing
source characteristics, etc.).
1.1Personnel
and equipment
A number of persons will be involved in this
research project:
·1 PhD
student, analysing, designing and realising a POF system based on mode group
diversity multiplexing techniques
·1 PhD
student, analysing, designing and realising an electronic signal processing
system for mode group diversity multiplexing
·0.3 fte
Technician, assisting the PhD students in realising the optical and electronic
system modules, integrating those system modules into a working mode group
diversity system setup, and testing its performance
·Two full
professors, each at 0.1 fte, guiding and acting as promotor for the two PhD
students
·One
assistant professor at 0.2 fte, and three associate professors, each at 0.1
fte, for daily supervision of the PhD students
Extensive
laboratory facilities are available to perform the POF system research. In
addition, however, in order to analyse the intensity distribution of the mode
fields at the end face of the multimode fibre (the so-called near field
pattern), a PC-based video frame grabber system (using a specific video
processing card and software) is needed to analyse the video images obtained
with a video camera mounted on a microscope. The costs of such a system are
about 14 kEuro.
1.1Time
planning and task allocation
The proposed duration of the project is 4 years,
involving two complete PhD tasks.
An overview of the task allocation to the project
participants and of the time planning is given in Table 2.
2.2 Available infrastructure
The proposed research activities in multimode polymer optical fibre systems are embedded in the COBRA research institute, and are complementary to a number of running POF research activities:
· Gigabit Ethernet data transport in graded-index POF links, in cooperation with Asahi Glass (Japan) providing latest graded-index POF samples
· Radio over POF, part of the B4 research project Broadband Radio @ Hand, executed in cooperation with KPN Research, Philips Research, and Agere Systems
The laboratory of the ECO group is well-equipped for optical system research, with advanced optical measurement setups.
The infrastructure of the group SPS is focussedfocused on
signal processing and coding.
· The group has extensive knowledge on signal-processors.
· There is a powerful workstation (HP9000, 4 CPU’s, 2Gbyte Ram) available for simulations of coding and signaling techniques.
· Facilities for programming Xilinx FPGA-s are available, and FPGA expertise has been built up.
2.3 Related research
Various groups worldwide are investigating data transmission over POF links. In Table 1, an overview is given of the record transmission capacities achieved up to now. The highest data rate over a single wavelength channel stands at 11 Gbit/s. Multi-channel techniques up to now deployed usually wavelength multiplexing; the record bitrate*length product is held by the ECO group at Eindhoven University of Technology, and stands at 2.28 Gbit×km/s. The record distance bridged so far is almost 1 km GI-POF with a Gigabit Ethernet data signal (1.25 Gbit/s), also achieved by the ECO group.
Table 11111 Perfluorinated graded-index
POF transmission records
|
Year |
Bitrate (Gbit/s) |
Distance (m) |
Wavelength
(nm) |
Organisation |
|
1997 |
2.5 |
200 |
1300 |
Fujitsu |
|
1998 |
5 |
200 |
1310 |
Eindh. Univ. |
|
1998 |
2.5 |
300 |
645 |
Eindh. Univ. |
|
1998 |
2.5 |
550 |
1310 |
Eindh. Univ. |
|
1999 |
2.5 |
550 |
840 |
Eindh. Univ. |
|
1999 |
11 |
100 |
1300 |
Lucent |
|
1999 |
7 |
80 |
950 |
Ulm Univ. |
|
1999 |
3l x 2.5 |
200 |
645, 840, 1300 |
Eindh. Univ. |
|
2000 |
2l x 2.5 |
456 |
840, 1300 |
Eindh. Univ. |
|
2001 |
1.25 |
990 |
840 |
Eindh. Univ. |
To our knowledge, up to now only one literature reference [7] has reported on mode group diversity multiplexing: Stuart et al. of Bell Labs showed in very early experiments based on the space-time coding BLAST techniques the basic feasibility with 2 transmitters and 2 detectors, each handling a bitrate of only 1 Mbit/s and where the data processing was done off-line in a PC. According to [6], selective mode launching may considerably enhance the bandwidth of POF links, up to a factor of 4; although it was not specifically mentioned in this reference, this further underlines the importance to combine selective mode launching with mode group diversity multiplexing for capacity enhancement.
For the related research on space-time processing we refer to the references [14][15][16]. Other than the one of Stuart et al. mentioned above, a literature search did not reveal activities closer related to coding for multi-mode diversity multiplexing than space-time coding
1.1
Urgency
The field of in-house networking is moving fast;
the demand for more bandwidth, for a wider range of more tailor-made services
(mostly IP-based) and for quality-of-service differentiation is growing at
increasing pace. Presently, different in-house infrastructures are deployed,
optimised for delivery of a limited set of services. No interworking between
these infrastructures exists, and continued development of dedicated in-house
networks will lead to quickly diverging gaps between these infrastructures and
to severe market segmentation.
Therefore solutions to obtain convergence of these
network infrastructures into a single versatile broadband infrastructure are
necessary. Such a convergence would also considerably ease upgrading and
maintenance of the in-house network, and facilitate the introduction of new
services. Thus solutions need to be explored on short notice for a universal
transparent broadband in-house infrastructure capable of hosting a wide variety
of services with different quality-of-service demands.
The proposed mode group diversity multiplexing
approach on multimode (polymer) optical fibre can provide such a universal
broadband in-house infrastructure, capable of integrating independently various
services with different characteristics. In the Netherlands, a number of
industries are active in in-house networking (Philips, Lucent Technologies,
KPN, Draka Fibre, Ericsson, KPN, …), and
the experience to be gained in this project can be very beneficial for making
their strategic urgent choices for product development. It will also enhance
the knowledge base in this important field in the Netherlands, foster talents
for the Dutch telecommunication industry, and strengthen their position as part
of their international company context.
3 Utilisation plan
3.1 Practical challenges and proposed solution
The main goal of the project is to
architect, design and validate a universal future-proof in-house network architecture
for seamless transparent communication within the residential
area, between the various in-house user terminals. Through a residential
gateway (as studied in complementary other projects executed by the proposers
at TU/e), this universal in-house network can flexibly interact with the
various access networks.
In the residential area, one may discern four major groups of services:
· Video (TV, VCR, DVD, videoconferencing, video streaming, etc.)
· Audio (sound, radio, voice)
· Data (PC, peripherals, internet)
· Control (of home appliances, domotica)
Today,
these services are run over a wide variety of mutually incompatible in-house
networks (coaxial cable, twisted copper pairs, power line, wirelessly)
which), which are not interconnected, and are
using different signal formats and service protocols. Moreover, user services
are often tied to specific locations, where specific network connections are
installed, thus hampering user mobility. As a solution to reduce the network
variety and increase the ease of use for the customer, the project proposes to
perform research into a new common
broadband in-house network infrastructure, based op Multimode (Polymer)
Optical Fibre. By means of the proposed mode group diversity multiplexing,
signal transparency and high bandwidth is offered in a single fibre
infrastructure, which enables high capacity signal transport as well as
integration of multiple services into a single infrastructure with universal
access at all network outlets. Thus significantly more bandwidth, flexibility
and mobility for services is offered to the user.
Complementary to this proposed research on a common in-house network architecture, concepts for a standardised modular Residential Gateway capable of handling and interconnecting all services in the residential area (i.e., in-house and access networks) are being investigated in other projects led by the proposers. Such a gateway will provide a future-proof, scalable, modular and universal solution for coupling access and in-house networks.
1.1
Identification of user group
Philips Research (group of dr.ir. P. Hooijmans,
located in Eindhoven) is interested in the mode group diversity multiplexing
for broadband multi-service in-home networks. They are willing to contribute
some integrated multi-photodiode arrays (so-called PDICs, which are being
researched for high-speed reading of data from optical disks), which can be
deployed in the multiple-receivers architecture of the proposed mode group
diversity system. However, it has to be verified with Philips Semiconductors,
whether these PDICs can be made available.
Draka Fibre Technologies (group of dr.ir. P. Matthijsse,
located in Eindhoven) is also interested in the proposed work, and is willing
to contribute samples of multimode silica fibre for experiments, and assessing
the potential of these fibres next to the polymer optical fibre.
Lucent Technologies Nederland (Advanced
Technologies – Bell Laboratories group, dr.ir. J.S. Wellen) is interested in
the application potential of mode group diversity multiplexing in in-building
networks, in particular for integration of multiple services in a single
in-building infrastructure, and how it will interact with the access network.
KPN Research (dr. F. den Hartog, group leader Home
Networking) regards the proposed technique as a very interesting one to solve
the bandwidth bottleneck and to enable the delivery of multiple services with
different transport requirements.
The above-mentioned parties are interested to
participate in the User group of the proposed projects; support letters have
been received (see attachments to this proposal).
For many years, there is also an excellent
cooperation with Asahi Glass (Japan) which), which provides
state-of-the-art graded-index polymer optical fibres to the ECO group. In
return, they are getting feedback on the system experiments in which the
quality and application potential of their fibres are assessed.
3.2 Implementation plans
Key criteria for in-house networks are ease of use and low cost.
The proposed mode group diversity multiplexing technique enables to integrate a variety of broadband services with widely ranging protocol requirements in a single transparent infrastructure, thus offering ease of use by providing easy upgrading with new services, easy accessibility throughout the house, and terminal mobility. Moreover, the polymer optical fibre can be easily installed due to its large flexibility and ductility.
To implement the system at low cost,
integration of the optical and electrical functions is essential. An integrated
array of vertical cavity surface emitting lasers (VCSELs) can advantageously be
deployed for launching the different mode groups at the transmitter end; it may
be coupled easily by means of a simple lens system, or even by butt-joining to
the large-core multimode (polymer) optical fibre. At the receiving end, an array of integrated
photodetectors can be easily coupled to the fibre’s large core, and
perform the spatially resolved detection of the mode groups.
The electrical signal
processing will be implemented first in Field Programmable Gate Arrays
(FPGA-s), offering design flexibility. After having demonstrated the technical
feasibility, the functions can be more cost-effectively integrated in
Application-specific IC-s (ASIC-s). of the
multiple lasers at the transmitting end can be
1.1
Past performance
The Electro-Optical Communication Systems (ECO) group is part
of the Dept. of Electrical Engineering, Eindhoven University of Technology. It
is performing research addressing the whole spectrum of telecommunication
infrastructures, ranging from ultra-high-speed long-reach core networks via
metropolitan and access networks to broadband residential networks. Optical
fibre is the prevailing medium, complemented with other first-mile and
residential area technologies such as VDSL over twisted copper pairs,
next-generation mobile and wireless LAN, and Gigabit Ethernet via polymer
fibres. Also universal modular residential gateway concepts are explored, which
support different service planes (video, audio/voice, data and control) and
allow interaction between different mutually incompatible in-house network
technologies, and provide a gateway to the outside world using various access
network technologies. Based on the insights gained, applications and
demonstrators are developed. Linux is used for its open structure and rich
network support. Within the ECO group, a sizeable group of students called
SpaceLabs (Society Pursuing Achievements in Communications Embracing Linux
Architectural BasiS, see www.spacelabs.nl , is
actively experimenting with novel broadband services.
The ECO
group is a key member of the National Research Centre Photonics, within the
research institute COBRA.
Within the
Brabant BreedBand (B4) alliance between the TU/e and a number of industries
(see www.brabantbreedband.nl ) , the
taskforce “Broadband TU/e Campus Network” has been formed to explore
implementation of broadband experimental networks and to facilitate
experiments. The most important users of the B4 network will at first be
students from the Electrical Engineering and Information Technology Faculty.
This population, as they have the right background, can contribute to the
development of advanced networking technologies by testing those (loading the
network) and giving feedback. In a next stage, these experiments can be
extended to groups of students on the entire campus as the accent shifts to
applications. Especially, video-enabled collaboration and digital learning
environments can be tested. Eindhoven University of Technology is leading in
Europe concerning computer-aided education. The general policy at TU/e is to
equip students with notebook computers. Presently, several thousands of
students already use the ICT services at the campus intensively. The ultimate
situation envisioned is a TU/e campus network with a polymer optical fiberfibre (POF)
fixed backbone infrastructure with broadband wireless extensions covering the
whole campus. The Wireless Classroom is the first pilot in the B4 framework.
The ECO
group has a long-standing experience in national and European research
projects; it participated/participates in the European Commission-funded
projects RACE-MUNDI, RACE-TOBASCO, ACTS-BLISS, ACTS-Upgrade, ACTS-Fleximacs, ACTS-APEX,
ESPRIT-LOCOM, IST-METEOR, IST-FASHION, and IST-STOLAS, and several B4 projects
running under the BTS programme (Broadband Radio@Hand, RETINA,
Residential Gateway Environment).
The Signal Processing Systems (SPS) group is part
of the Department of Electrical Engineering of Eindhoven University of Technology.
Together with the Control Systems group, it is clustered in the sub-department
"measurement and control systems" (MBS). Signal
processing technology is a cornerstone of the electronics industry, and is
central to many exponents of modern society, such as wired and wireless
communication systems, digital audio and TV systems, multimedia PCs, advanced
medical equipment, and even cars. The SPS group contributes to this technology
through a balanced mixture of fundamental and systems-oriented research.
The fundamental research aims at novel theories and
algorithmic building blocks, and is focused on areas that are expected to play
a key enabling role in future signal-processing systems, such as:
·
signal transforms and filter
banks,
·
adaptive array signal
processing, and
·
semantic signal processing.
The systems-oriented research integrates existing
signal-processing theory and building blocks into innovative algorithms, system architectures and
implementations. This research is focused on:
·
acoustic and audio
communication,
·
digital communication, and
·
medical signal processing.
These contexts are selected because they are
technically challenging, relevant to society and industry, and synergetic with
the fundamental research. Moreover, in-depth contextual know-how is available
through strategic partnerships, e.g. with Philips Research Laboratories in
Eindhoven, the Data Storage Institute in Singapore, and the Catharina Hospital
in Eindhoven. Such partnerships are essential for the quality and relevance of
the research, and in the longer run for its commercial potential.
Together with the Control Systems group, the SPS
group forms the newly established departmental research spearhead on
"Adaptive Systems".
Curricula Vitae of proposers
Ton Koonen (a.m.j.koonen@tue.nl) was born
in Oss, The Netherlands, in 1954. He is a full professor in Broadband
Communication Networks at the COBRA Institute, in the Department of Electrical
Engineering at Eindhoven University of Technology, the Netherlands. Before
joining the university, he served from 1979 to 2000 as member of technical
staff and as technical manager at Bell Laboratories, Lucent Technologies in the
Netherlands, where he led a group working on applied research in fibre-optic
broadband access networks. From 1991 to 2000 he also was at the University of
Twente in the Netherlands as a part-time professor in Photonic Networks, part
of the Telematica Systems and Services group, and affiliated to the CTIT, where
he worked closely together with prof. Ignas Niemegeers. He has initiated and
managed various projects in this field in the European ACTS and IST programs,
and is currently managing the IST project STOLAS on optical packet routed
networks. He is also managing the BraBant BreedBand (B4) alliance, a
pre-competitive research co-operation on broadband network techniques and
applications between the university and various industrial partners. He has (co‑)authored
over 65 papers on optical fibre communication. In 1999, he received the Bell
Labs Fellow award “for outstanding contributions to high-speed transmission
systems and for advancing optical technologies to support the convergence of
access networks”. He is also a Senior Member of IEEE.
Henricus P.A. van den Boom was born in Eindhoven, The
Netherlands in 1955. He received the degree of Elektrotechnisch
Ingenieur from the Eindhoven University of Technology, Eindhoven The
Netherlands, in 1984. Since then he has been an Assistant Professor at the
Electro-Optical Communications group of the Department of Electrical
Engineering of the same university. He lectures in basic telecommunication
theory and optoelectronic communication systems and networks. He has
(co‑)authored over 50 papers on optical fibre communication. He has been
involved in research on coherent optical communication systems, optical
cross-connected networks and broadband communications in Hybrid Fibrer Coax
networks. Currently he is working on Residential Area networks and Polymer
Optical Fibrer systems
and networks.
Frans M.
Huijskens was born in Oudenbosch, The
Netherlands, in 1958. He graduated in applied physics from the Technical
College of Dordrecht in 1979. From 1981 to 1984, he was an Electronic Test
Engineer at Siemens Gammasonics. In 1985, he joined the Electro-Optical
Communications Group of Eindhoven University of Technology, Eindhoven, The
Netherlands. His work involved research on passive fiberfibre couplers
and support on projects concerning phase and polarization diversity coherent
systems. He contributed in the development of optical cross-connect
demonstrators and in packaging technologies of optical integrated devices
within the ACTS APEX project. Currently, he is involved in experimental work on
multimode fibre systems.
Jan W.M. Bergmans (IEEE SM'91) was born in Tilburg, The Netherlands,
on August 15, 1957. He received the degree of Elektrotechnisch Ingenieur, cum
laude, in 1982 and the Ph.D. degree in 1987, both from Eindhoven University of
Technology. From 1982 to 1999 he was with Philips Research Laboratories,
Eindhoven, The Netherlands, working on signal-processing techniques and
IC-architectures for digital transmission and recording systems. In 1988 and
1989 he was exchange researcher at Hitachi Central Research Labs, Tokyo, Japan.
Since 1999 he is professor and chairman of the signal processing systems group
at Eindhoven University of Technology. Since 1998 he is advisor to the Data
Storage Institute in Singapore, and since 2000 to Philips Research Laboratories
in Eindhoven, The Netherlands.
He has published extensively
in refereed journals, has authored a book (`Digital Baseband Transmisison and
Recording, Kluwer Academic Publishers, 1996, 652 pp.), and holds around 30 U.S.
patents. From 1986 to 1992 he has been a member of the board of the board of
Dutch Electronics and Radio Society and the Dutch URSI Committee. Since 1999 he
is Treasurer of the IEEE Benelux Section.
Frans M.J. Willems was born in Stein, The Netherlands, in 1954. He
received the M.S. degree in electrical engineering from Eindhoven University of
Technology, Eindhoven, The Netherlands, and the Ph.D. degree from the Catholic
University of Louvain, Louvain, Belgium, in 1979 and 1982 respectively. From
1979 to 1982 he was a research assistant at the Catholic University of Louvain.
Since 1982, he is a staff member at the Electrical Engineering Department of
Eindhoven University of Technology. His research contributions are in the areas
of multi-user information theory and noiseless source coding. Dr. Willems
received the Marconi Young Scientist Award in 1982. From 1988 to 1990, he
served as Associate Editor for Shannon Theory for the IEEE Transactions on
Information Theory. He is co-recipient of the 1996 IEEE Information Theory
Society Paper Award. From 1998 to 2000 he was a member of the Board of
Governors of the IEEE Information Theory Society. Since 2001 he is Associate
Editor for the European Transactions on Telecommunications. He is an advisor
for Philips Research Laboratories Eindhoven.
Martin J. Bastiaans was born in Helmond, the Netherlands, on January
18, 1947. He received the 'ingenieur' degree (= M.Sc. degree) in Electrical
Engineering (with honours) and the Ph.D. degree in Technical Sciences from the
Technische Universiteit Eindhoven (Eindhoven University of Technology),
Eindhoven, the Netherlands, in 1969 and 1983, respectively.
Since 1969
he has been 'wetenschappelijk medewerker' (= Assistant Professor) and since
1985 'universitair hoofddocent' (= Associate Professor) with the Department of Electrical Engineering, Technische Universiteit Eindhoven, in the Signal Processing Systems Group, where he
teaches electrical circuit theory,
digital signal processing, and Fourier optics and holography. His research
covers different aspects in the general field of signal and system theory, and
includes a signal-theoretical approach of all kinds of problems that arise in
Fourier optics, such as partial coherence, computer holography, optical signal
and image processing, and optical computing. His main current research interest
is in describing signals by means of a local frequency spectrum (for instance,
the Wigner distribution function, the sliding-window-spectrum, Gabor's signal
expansion, etc.).
He served
for nine years as a member of the Research Committee of the Department of
Electrical Engineering in 1978/1982 and 1993/1998, and for almost four years as
one of the two vice-deans of this Department in 1984/1986 and 1991/1992. In
these functions, and particularly in the function of vice-dean, he was
responsible for the departmental research. Since 1 June 1998, he has been one
of the three members of the Board of the Department of Electrical Engineering,
and since the beginning of 2001, he serves as the Department's coordinator for International
Educational Activities. He is the Department's
representative in the University's Research
Information System Committee, and he is one of
the Department's representatives in the 'Stichting PATO', the Dutch Organisation of
Post-Academic Technical Studies. He is also the Counselor of the IEEE Student Branch Eindhoven and a
member of the IEEE Region 8 Student Activities Committee, where he
co-ordinates the Student Paper Contest.
Dr.
Bastiaans is a Fellow of the Optical Society of America - "in recognition of
distinguished service in the advancement of optics, particularly for pioneering
contributions to the description of optical signals using local frequency
spectrum concepts" - and a senior member of the Institute of Electrical and
Electronics Engineers. He has published about 100 papers in international scientific journals and
proceedings of scientific conferences.
Piet
Sommen received the Ingenieur
degree in electrical Engineering from Delft University of Technology in 1981
and his Ph.D. from Eindhoven University of Technology in 1992. From 1981 to
1989 he was with Philips Research Laboratories, Eindhoven and since 1989, with
the faculty of Electrical Engineering at Eindhoven University of Technology,
where he is currently an Associate Professor. Dr. Sommen is involved in
internal and external courses, all dealing with different Basic and Advanced
Signal Processing topics. His main field of research is in Adaptive Array
Signal Processing, with applications in Acoustic Communication Systems.
Dr. Sommen is a member of
the ProRISC board, Vice President of the IEEE Benelux Signal Processing Chapter
and officer of the Administrative Committee of EURASIP. Besides, he is Editor
of EURASIP's Journal of Applied Signal Processing. For this journal, he will
also serve as Guest Editor of a special issue on "Signal Processing for
Acoustic Communication Systems" that is scheduled for the end of 2003.
1
Intellectual Property
management
1.1
Contracts
No contracts related to the work proposed here
exist.
1.1
Patents
No patents regarding the proposed work have been
applied for by the proposers.
1
Budget
1.1
Personnel
We apply for funding of two new PhD student
positions, each for a period of 4 years.
Furthermore for funding of effort of permanent
staff, involved with guidance and daily supervisiosupervision, and
technical support.
1.1
Materials and travel
Material
costs:
PhD student 1: 12
kEuro per year, for optical components (surface emitting laser diodes,
photodiodes, optical connectors, optical power splitters and combiners, IC-s
for optical transmitters and receivers, etc.)
PhD student 2: 12
kEuro per year, for electronic components (such as FPGAs, digital signal
processors, analog/digital converters) and hardware for PC controllers
Travel:
5000 Euro per year per PhD student and his/her
supervisor, for visiting national and international conferences and workshops
1.1
Investments
To characterise the near field patterns of the
light distribution at the fibre output, which is a key issue to investigate for
distinguishing the mode groups, a PC-based video frame grabber system is needed
to analyse the video images obtained with a video camera mounted on a
microscope. This system uses a specific video processing card and software, and
costs about 14 kEuro.
1.1
Support by members of the
user group
-
multimode silica fibres from
Draka Fibre Technologies (dr.ir. Piet Matthijsse)
-
integrated fast photodiode
array (PDIC) from Philips Research (dr.ir. Pieter Hooijmans)
The members of the user
group are also interested to participate in regular meetings about the
project’s progress.
1.1
Budget summary
(amounts
in Euro-s)
New investments:
14 kEuro for video frame
grabbing card and software (for characterisation on near-field intensity
patterns on fibre endface)
4 Literature references
[[1]]
Ton Koonen, Henrie van den Boom,
Idelfonso Tafur Monroy, Giok-Djan Khoe, “Broadband data communication
techniques in POF-based networks”, Proc. of ECOC 2001, Amsterdam, Sep. 30 –
Oct. 4, 2001, Vol. 1, pp. 14-15
[[2]]
H.P.A. van den Boom, W. Li, P.K. van
Bennekom, I. Tafur Monroy and G. D. Khoe, “High Capacity Transmission over
Polymer Optical Fiber”, IEEE Journal on Selected Topics in Quantum Electronics,
Vol. 7, May/June, 2001
[[3]]
H.P.A. v.d. Boom, T. Onishi, Y.
Watanabe, Y. Tsukamoto, P.K. van Bennekom, L.J.P. Niessen, G.D. Khoe, A.M.J.
Koonen, "Gigabit Ethernet transmission over nearly 1 km GIPOF using an 840
nm VCSEL and a silicon APD", Proc. of POF 2001, Amsterdam, Sep. 27-30,
2001, pp. 207-211
[[4]] G.S. Yabre, “Comprehensive theory of dispersion in graded index optical fibers”, J. Lightwave Technol. 18, nr. 2, 2000, pp. 166-177
[[5]] G.S. Yabre, “Influence of core diameter on the 3-dB bandwidth of graded-index optical fibers”, J. Lightwave Technol. 18, nr. 5, 2000, pp. 668-676
[[6]] L. Raddatz, I.H. White, D.G. Cunningham,
M.C. Nowell, "An experimental and theoretical study of the offset launch
technique for the enhancement of the bandwidth of multimode fibre links",
IEEE J. of Lightw. Techn., Mar. 1998, pp. 324-331
[[7]] H.R. Stuart, "Dispersive multiplexing
in multimode fiber", Proc. of OFC 2000, Baltimore, Mar. 7-10, 2000
[[8]] T. Ishigure, Y. Koike, “Theoretical understanding of POF”, Proc. of 10th Intern. Plastic Optical Fibres Conf. (POF 2001), Amsterdam, Sep. 27-30, 2001, pp. 293-300
[[9]] M.J. Bastiaans, "Application of the Wigner distribution function to partially coherent light," J. Opt. Soc. Am., Opt. Image Sci., 1986, pp. 1227-1238
[[10]] M.J. Bastiaans, "Application of the Wigner distribution function in optics," in The Wigner Distribution - Theory and Applications in Signal Processing, ISBN 0-444-88856-X, ed. W. Mecklenbräuker; F. Hlawatsch; Elsevier Science, Amsterdam, Netherlands, 1997, pp. 375-426
[[11]] M.J. Bastiaans, "Gabor's signal expansion in optics," in Gabor Analysis and Algorithms: Theory and Applications, ISBN 0-8176-3959-4, ed. H.G. Feichtinger; T. Strohmer; Birkhäuser, Boston, MA, 1998, pp. 427-451
[[12]] A.M.J. Koonen, “Bit-error-rate degradation in a multimode fiber optic transmission link due to modal noise”, IEEE J. on Selected Areas in Comm., Dec. 1986, pp. 1515-1522
[[13]] E. Teletar, “Capacity of multi-antenna Gaussian channels,” Tech. Memo., AT&T Bell Labs, June 1995
[[14]]
G.J. Foshini and M.J. Gans, “On limits
of wireless communications in a fading environment when using multiple
antennas,” Wireless Commun. Mag., vol. 6, pp. 311-335, Mar. 1998