Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
Increased Capacity Bidirectional DWDM Network Architecture
With Frequency Stacking System
FIELD OF THE INVENTION
The present invention relates generally to a cable television hybrid-fiber-
coax
(CATV HFC) architecture that provides increased capacity in the reverse path
of the
network. More particularly, it describes an architecture that incorporates
multiplexing,
both optical multiplexing, using dense wave division multiplexing (DWDM), and
RF
multiplexing, using frequency stacking, so as to increase the efficiency of
the return
path.
BACKGROUND OF THE INVENTION
The evolution of the traditional CATV HFC network into a two-way interactive
data communications platform (including cable modems and IP telephony over
cable),
together with the move toward headend consolidation, has led to a need for
more
bandwidth in both the forward and reverse paths.
Specifically, in a typical CATV plant today, downstream content occupies the
50-870 MHz frequency partition of the network. The return path signals are
relegated
to the 5-42 MHz frequencies (of course those skilled in the art will
appreciate that
although the typical reverse path frequency band in the United States is 5-42
MHz,
overseas the range varies and may be 5-85 MHz - the concepts discussed herein
are not
to be interpreted as being limited to the current US range). Given the
asymmetrical
nature of the frequency bands used it is very likely that the reverse path
traffic will be
the first to be constrained.
DWDM systems have been deployed to provide segmentation and increased
bandwidth. However, by itself, DWDM can add impairment that was not present in
the
system earlier. In addition, multiplexing in the RF domain using frequency
stacking
has been used in the reverse path passband to increase bandwidth efficiency.
That is,
the implementation of a frequency stacking system expands the return bandwidth
per
home passes and allows for larger node sizes, thereby reducing the overall
system
costs. Again, however, by itself frequency stacking can add impairment to the
system.
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
The present invention is therefore directed to the problem of developing an
architecture that meets the current needs of multiplexed analog and digital
systems,
segments the forward signal to address individual subscribers and increases bi-
directional capacity without the use of additional fiber.
SUMMARY OF THE INVENTION
The present invention provides a hybrid DWDM with frequency stacking
architecture that solves the current needs of muliplexed analog and digital
systems to
provide maximum capacity based on two-way interactive data communications.
According to one embodiment of the present invention, a CATV architecture
provides increased capacity in the reverse path network for two-way cable
communication and includes a plurality of optical-to-electrical conversion
nodes, a
primary/secondary headend, and a master headend. The primary/secondary headend
interconnects the optical-to-electrical conversion nodes and master headend.
The nodes
and primary/secondary headend together includes an upconverter for receiving a
plurality of RF reverse path passbands from a plurality of coax legs and
upconverting
the return passbands to different passbands, a plurality of DWDM transmitters,
each
transmitter having an output on the ITU grid and transmitting a concentration
of
discrete passbands and a DWDM multiplexes, receiving a signal from each of the
plurality of DWDM transmitters and optically multiplexing the DWDM
transmitters on
a single fiber, the multiplexed signals being routed to the master headend.
The master
headend includes a DWDM demultiplexer for demultiplexing the received signals
from
the DWDM multiplexes into individual wavelengths, a plurality of block
conversion
receivers (BCRs) for receiving the individual wavelengths and converting the
signals
into composite RF signals and a plurality of block downconverters (BCDs) for
receiving the composite RF signals from the BCRs and converting the signals
into
individual RF signals. The individual RF signals output from the plurality of
BCDs
correspond to the plurality of coax legs at each optical-to-electrical
conversion node.
Another aspect of the invention incorporates an optical amplifier, at the
primary/secondary headend, that amplifies the multiplexed signals output from
the
DWDM multiplexes before being routed to the master headend. In a particular
embodiment, the optical amplifier may be an erbium-doped fiber amplifier
(EDFA).
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
Still yet another aspect of the invention includes using Code-Division-
Multiple-
Access (CDMA), Frequency-Division-Multiple-Access (FDMA), Time-Division-
Multiple-Access (TDMA), or any combination thereof, to allow the transport
link's
available capacity, defined by channel parameters, to be achieved.
Another embodiment of the invention is directed to a method for increasing
capacity in the reverse path of a two-way cable communication architecture,
the
architecture having a plurality of optical-to-electrical conversion nodes, a
master
headend, and a primary/secondary headend interconnecting the nodes and the
master
headend. The steps of the method are as follows: receiving a plurality of RF
reverse
path passbands from a plurality of coax legs and upconverting the return
passbands to
different passbands, transmitting a concentration of discrete passbands using
a plurality
of DWDM transmitters, each transmitter having an output on the ITU grid,
optically
multiplexing the signals received from the DWDM transmitters, using a DWDM
multiplexer, on a single fiber, routing the multiplexed signals to the master
headend,
demultiplexing the received signals into individual wavelengths, receiving the
individual wavelengths and converting the signals into composite RF signals
and
receiving the composite RF signals and converting the signals into individual
RF
signals.
BRIEF DESCRIPTION OF THE DRAWINGS
The above-mentioned and other objects, features and advantages of the present
invention will become more apparent from the following detailed description
when
taken in conjunction with the accompanying drawings, wherein:
FIG. 1 depicts a typical tree-and-branch configuration of a hybrid Fiber/Coax
(HFC) cable TV network architecture;
FIG 2 depicts a DWDM subcarrier multiplexed network architecture;
FIG 3 depicts a typical architecture for a DWDM overlay of a standard CATV
distribution system;
FIG 4 depicts a block diagram of a typical node configuration;
FIG 5 depicts a typical frequency stacking system (FSS) block diagram;
FIG 6 depicts a fist embodiment of an architecture according to the present
invention combining DWDM and FSS technologies;
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
FIG 7 depicts a first embodiment of an architecture according to the present
invention combining DWDM and FSS technologies;
FIG 8 depicts a second embodiment of an architecture according to the present
invention combining DWDM and FSS technologies;
FIG 9 depicts a third embodiment of an architecture according to the present
invention combining DWDM and FSS technologies;
FIG 10 depicts a composite RF spectrum of four upconverted return-path
frequency blocks; and
FIG 11 depicts the bandwidth expansion provided by a combined frequency
stacking and DWDM return path system.
DETAILED DESCRIPTION
Typical CATV systems were almost exclusively designed for "one-way
transmission" from a headend to the home. Return path implementations were
typically lightly loaded and used primarily for low-speed communications with
terminals or set top boxes. The recent use of DWDM in CATV was motivated by
the
need to significantly increase the bi-directional capacity as well as
transmission access
speed without adding additional fiber
The present invention is directed to a hybrid DWDM/FSS CATV architecture
that provides increased capacity in the reverse path network. Consequently,
the present
invention has significantly improved the ability to handle two-way interactive
multimedia communications in existing CATV systems while eliminating the need
for
the use of additional fiber.
The descriptions of the specific embodiments of the invention follow the brief
discussion of DWDM and frequency stacking systems below.
I. DWDM
DWDM systems in CATV today are used exclusively in the 1550 nm optical
window (this wavelength window is attractive primarily due to the low fiber
loss,
approximately 0.22 dB/km at the 1550-nm wavelength, and the use of erbium
doped
fiber amplifiers (EDFA) to take advantage of the low loss). The wavelengths
that
comprise the ITU grid are actually a set of predefined frequencies, such as
the set
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
spaced at 100 GHz, from which wavelengths can be derived. The wavelength
spacing
is approximately 0.8 nm, and the range of wavelengths covers the EDFA band,
from
about 1530 to 1570 nm. Of course, not all the wavelengths need to be used in
any
given system and commercial products are available at 100, 200, and 400 GHz
spacings, with. a variety of individual product offerings. In a preferred
embodiment of
the invention to be discussed in detail below, the spacing chosen is 200 GHz,
and it is
this closeness of wavelengths that make the system "dense" (this is to
distinguish the
DWDM systems from some existing CATV systems which use a combination of 1310
nm and 1550 nm wavelengths in a WDM arrangement). In CATV jargon, the RF
signals transmitted over the single-mode fiber (SMF) cables using, for
example,
DWDM transmitters (such as transmitter 320 in Figure 3 described below), are
digital
(but of course could also be analog or a combination of digital and analog)
using, for
example, QAM modulation. The QAM channels are subcarrier multiplexed onto a
particular optical wavelength (the terms QAM, digital, targeted services, or
DWDM
signals are frequently used interchangeably).
Generally, Figure 1 shows a conventional hybrid fiber/coax (HFC) cable TV
network architecture. As shown, the signals from the master headend 10 are
connected
via a "main" or "primary" fiber ring to primary/secondary headends ( 12a, 12b,
12c) or
the secondary "hubs" in a large metropolitan area (14a, 14b, 14c and 14d). The
signals
are transmitted over single-mode fiber (SMF) using, for example, 1550-nm
externally
modulated (EM) DFB laser transmitters. The composite signal may be, for
example, a
mixture of traditional broadcast analog signals with MPEG compressed digital
video.
At the primary and secondary headends, which may house Synchronous Optical
Network (SONET) equipment as well as Cable Modem Termination Systems
(CMTSs), routers, and servers for high-speed data, the optical signals may be
converted
to RF signals and then back to optical signals for transmission to various
fiber nodes
(16a, 16b, 16c and 16d) using, for example, 1310-nm DFB laser transmitters.
The coaxial portion of the network architecture illustrated in Figure 1,
consists
of, for example, RF amplifiers, taps, and coaxial cables, and spans from each
fiber node
( 16a-d) to the corresponding subscriber's home(s), where the digital set-top
box is
placed.
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
Figure 2 provides an exemplary illustration of a DWDM Subcarrier-
Multiplexed (SCM) network architecture for multiple AM/QAM channel transport.
In
this network architecture, the master cable-TV headend 10 is connected via a
main fiber
ring to the primary headends (12a, 12b, 12c) in a large metropolitan area. The
analog
and digital video programs at the master headend 10 are typically received via
satellite
and terrestrial broadcast as well as via local video servers (those skilled in
the art will
appreciate that the analog signals could also be "injected" or received at the
primary or
secondary headends). Ultra-high video trunking capacity is achieved by using
high-
density wavelength-division-multiplexing (DWDM) and demultiplexing with
cascaded
Erbium-Doped-Fiber-Amplifiers (EDFA's) (23a, 23b, 23c). In the primary fiber
ring,
multiple wavelengths in the 1550-nm band, transmitted with each wavelength a
mixture
of AM and digital video signals, are subcarrier multiplexed (SCM) at either
passband
traffic, using 64/256-QAM, or even at baseband traffic, such as OC-48. The
secondary
fiber rings connect the various primary headends to the secondary headends. In
a
secondary fiber ring, only a few wavelengths are being demultiplexed and
transmitted
using 1550-nm or 1310-nm laser transmitters with cascaded EDFA's. At the
secondary
headends, the 1550-nm based broadcast traffic can be switched to a 1310-nm
based
traffic for both narrowcasting and broadcasting services. In each fiber node,
the optical
signals transmitted downstream at different wavelengths are converted back to
electrical signals using optical receivers, and are transmitted to each
subscriber via the
coaxial cable plant.
In order to address the return path, Figure 3 shows a simplified master
headend
300 (corresponding to reference number 10 of Figures 1 and 2) -
primary/secondary
headends/hubs 330 (corresponding to reference numbers 12a, 12b, 12c and 14a,
14b,
14c and 14d, respectively, of Figures 1 and 2) -- and node 360 (corresponding
to node
reference numbers 16a-16d of Figures 1 and 2). This figure is a generic
architecture for
a DWDM overlay of a conventional hybrid fiber/coax (HFC) cable TV network
(note
that for exemplary reasons only, Figure 3 assumes that the optical network
remains in
the 1550nm window from the master headend to the node), and includes an
externally-
modulated analog transmitter source 305 and a externally modulated DWDM
transmitter 320 (of course, this does not have to be a DWDM externally
modulated
transmitter but may be a directly modulated transmitter). As shown in Figure
3, the
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
master headend is collectively 300, the primary/secondary headends/hubs are
collectively 330, and the nodes collectively 360 (individually 360a, 360b
etc.) (those
skilled in the art will appreciate that although illustrated as such, the ITU
transmitter
does not necessarily need to be co-located with the broadcast transmitter).
The DWDM transmitter 320, illustrated as being located at the master headend,
includes laser modules for providing the bias, temperature, pre-distortion
circuitry and
monitoring controls as well as the means of modulating the sources with the RF
content. That RF content is either the analog broadcast television signals or
the
targeted services QAM signal. The modulation techniques are either external
(using the
modified balanced-bridge Mach-Zehnder interferometer) or direct (using the
driving
current control of the laser directly).
The output of analog transmitter source 305 is optically amplified 307 to a
saturated level (for example, approximately +17 dBm), transmitted through 40
km of
standard (non-dispersion shifted) single-mode fiber (SMF) (again, the length
of the
SMF is provided solely as an exemplary value) to the primary/secondary
headend,
amplified again by an erbium doped fiber amplifier (EDFA) 335 and split by
optical
splitter 340 into a number of outputs that matches the number of targeted-
services
wavelengths.
After splitting, the analog signal is multiplexed with the QAM wavelengths in
an analog/digital coupler 350 and that composite signal is again split to
serve a number
of nodes 360 for which the given wavelength is targeted. In this generic
system, nodes
360 are "20 km" away from the primary/secondary headend and are connected
using
standard SMF (note that there may be multiple nodes targeted per wavelength,
especially in the early deployment stages when subscriber take rates are low
corresponding to a low bandwidth requirement per node).
Returning to Figure 3, it should be appreciated that the DWDM laser sources
are also externally modulated transmitters 320 in the example system, but
directly
modulated sources may also be used. Eight wavelengths are shown in the figure
and
are combined into a single fiber in a multiplexes 325 (with 200 GHz spacing, 8
wavelengths may be multiplexed). The multiplexes 325 (and demultiplexer 355
described below) components are used to combine the various ITU grid
wavelengths
through a low loss coupler and carry them on a single fiber (and demultiplexer
355
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
subsequently separates those wavelengths to place them onto individual
fibers). While
the SMF may of course be any length, the standard SMF is 40 km long, and may
be
distinct from the fiber carrying the analog signal, but may be in the same
optical fiber
cable bundle (each optical fiber cable bundle consisting of multiple optical
fibers).
After the 40-km, at the primary/secondary headend location, the combined
wavelengths
are amplified by EDFA 357 and are then demultiplexed 355 into separate fibers.
As
noted above, each targeted services wavelength is combined with one of the
split
analog signal outputs and distributed to nodes 360 through a single fiber
carrying both
the analog and digital signal. The fiber node 360 contains a receiver that
detects both
the analog and QAM signals for distribution through the RF plant beyond the
node.
In keeping with the drive towards a more symmetric network, the return path
illustrated in the DWDM overlay system of Figure 3 mirrors that of the
downstream
path. One exception to this mirroring occurs not so much in the single fiber
of the
DWDM system, but in the portion of the return from the node 360 to
primary/secondary headend 330. The return path is managed as a two-hop
process. In
the illustrated system, a temperature compensated 1310/1550 nm laser
(typically a
DFB) is in the node 360. The time and frequency division multiplexed RF
signals from
each home served by this node (e.g., 1000-1200 subscribers) drive the DWDM
laser
385. The optical output is sent over the link (illustrated as "20 km"), to the
primary/
secondary headend 330, where it is detected and amplified by a return receiver
380
before directly modulating an ITU grid DWDM laser transmitter 385. The laser
385 is
one of several which combine the entire return path into a DWDM set for
transmission
over the 40 km back to the master headend 300 and for subsequent processing.
Each of
the DWDM wavelengths may handle the return traffic from multiple nodes 360
using a
combination of time, frequency, or code division multiplexing.
As noted above, the network solution shown in Figure 3 assumes that the
optical
network remains in the 1 SSOnm window from the master headend to the node. If
however an existing system utilizes a re-transmission scheme at the
primary/secondary
headend, there remains the goal to preserve as much of this infrastructure as
possible.
Fortunately DWDM can still be used to provide the narrowcast overlay.
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
II. Freq_uency Stacking
In frequency stacking systems the 5-42 MHz return passband is block
upconverted or shifted to another frequency passband. This may be done in a
primary/secondary headend environment or, as we will also discuss herein, in
the field
located node. The main advantage of the implementation of a frequency stacking
system (FSS) is the expansion of the return bandwidth per home passed, which
allows
for larger node sizes, which in turn reduces the overall system costs (more
specifically
implementation of FSS provides an expansion in which the same number of users
can
use a higher speed, or, the system can have a greater number of users).
If we look at "typical" node configuration, shown in Figure 4, all the users
served by the node share the return path spectrum. If this were a 1200 home
passed
node, each home passed would have approximately 29 KHz of guaranteed
simultaneous
bandwidth (this assumes that the entire 35 MHz is available, and we can
dynamically
allocate the bandwidth). As Figure 4 illustrates, each of the coaxial busses,
RF Leg # 1,
1ZF Leg #2, etc., are 1ZF' combined into one stream.
Adding more transmitters combined with segmenting the 1RF paths within the
node may increase bandwidth. However, this approach has disadvantages. Beyond
adding one additional return transmitter in the node, which only doubles
capacity, fiber
availability issues may become the limiting factor. To achieve the same level
of
bandwidth per home passed as FSS three additional transmitters and fibers
would be
required.
An FSS approach utilizes upconversion in the node to create four passbands for
the return. In this approach each leg now has its own 35 MHz of space. The
four
passbands are 1RF stacked and sent to the return laser. Figures 5 and 6
illustrate this
arrangement.
As Figures 5 and 6 illustrate, there are four major components associated with
a
FSS system - an upconverter, a transmitter, a receiver and a downcoverter.
These
components would be common in function regardless of whether the application
is hub
or node based. Each of these components is briefly discussed below.
Frequency stacking begins with the upconverter 500. This device, simply put,
takes multiple return passbands and shifts them to other independent passbands
in the
spectrum while maintaining the information that resides in the original
passband. In the
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
implementation shown in Figures 5 and 6, each of the RF legs is upconverted to
different passbands within the 50-1100 MHz passband. A pilot carrier serves
two key
functions - first, it compensates for the range of link loss introduced by the
optical
network, and second, it is used by the downconverter to phaselock to the
upconverter
thus eliminating frequency offsets.
The transmitter used in this application is not a standard, band-limited,
return
path transmitter. In this implementation, a forward path transmitter 510,
designed to
operate in the 50-400 MHz passband, is used to transport the upconverted
signal from
upconverter 500.
The FSS receiver (BCR) 520 is also different than the normal return path
receiver (RPR 410 of Figure 4). Again chosen for the forward path, the
receiver 520
provides the composite RF output. Contained within this passband are the four
upconverted bands along with the pilot carrier. To recover the individual
bands, a
downconversion process is performed by downcoverter 530, which provides the
means
of returning the upconverted bands to their original 5-42 MHz spectrum. Using
the
pilot carrier for frequency synchronization, the block downconverter (BCD) 530
reverses the process initiated in the node and provides four independent 5-42
MHz
passbands, one for each of the upconverted bands. These outputs are then fed
to the
return splitting/combining network and eventually end at the individual
service
demodulators.
III. Combined DWDM and Frequency Stacking Systems
With the above descriptions of DWDM and FSS systems, Figures 7-9 illustrate,
respectively, first, second and third embodiments of a combined system
according to
the present invention. Each of the approaches work together so as to increase
the
efficiency of both the return distribution and return transport aspects of the
network and
allow the combined exemplary system to have thirty-two 5-42 MHz return bands
on a
single fiber. The main difference between the embodiments, as will be clear
from the
description below, is the location of the ITU grid transmitters and the
location of the
frequency stacking system (the network architecture of Figure 7 has the DWDM
transmitters at the primary/secondary headend, Figure 8 has the DWDM
transmitters at
0
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
the node and Figure 9 has both the frequency stacking system and the DWDM
transmitters at the primary/secondary headend).
In the first embodiment of Figure 7, ITU grid DWDM transmitters 750 are
located at the primary/secondary headend. As illustrated, this configuration
upconverts
the return path signals, with upconverter 766, at the node location
(collectively the
nodes are 760). Transmitted back to the primary/secondary headend via the
optical
distribution network they are received by the forward path block conversion
receiver
(BCR) 735.
Unlike a standard FSS network, which then forwards the RF output from the
receiver to a downcoverter, in the first embodiment of the invention, the RF
output is
routed to a DWDM transmitter 750 which has an output wavelength on the ITU
grid.
The specific embodiment shown in Figure 7 has a concentration of four discrete
5-42
MHz passbands on each of these transmitters (of course, those skilled in the
art will
appreciate that the number of passbands on each of the transmitters can be
greater than
the illustrative "4" passbands shown and in fact, is limited based only on how
much the
laser can handle).
Using 200 GIIz spacing, for example, the configuration of Figure 7 can
optically multiplex (multiplexes 760) eight of the transmitters 750, each with
its own
different ITU grid wavelength, on a single fiber, providing 32 discrete 5-42
MHz
passbands (1.12 GHz) on the single fiber, thereby clearly illustrating how the
combination of FSS and DWDM significantly increases the reverse path traffic
capacity. The signals are then routed to the headend (note that depending on
the
distances involved, and requirements such as redundancy, optical amplifiers
may be
required to meet the input requirements of the headend receivers).
At the headend the optical signals are demultiplexed by demultiplexer 770 (in
the exemplary demultiplexer shown, into the eight wavelengths). Individual
wavelengths are routed to receivers (one for each wavelength) BCRs 780 which
are the
same type as those used to receive the frequency-stacked multiplex at the
primary/secondary headend. At this point the FSS system may be completed by
routing
the composite RF signals from the BCRs 780 to downconverters 790. The four, 5-
42
MHz, RF outputs from the downconverter 790 correspond to the four coaxial legs
n
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
coming into the field node 760, and may be routed to the various return path
application
receivers.
It is important to note however that there is no inherent requirement to use
downconversion from a communications system standpoint, but rather is
dependent
upon the hardware implementations. The primary/secondary or master headend
receiver implementations may anticipate an RF signal in the 5-42 MHz range and
may
be designed to frequency convert this range of spectrum for subsequent
processing,
thereby requiring a downconverter component. However, implementing these
receivers
instead with an input bandwidth capability that encompasses the FSS spectrum
would
eliminate the need for the downconverter component. For example, instead of
having
two downconversion components before processing (one external to the receiver,
and
one in the receiver), a more efficient implementation could accomplish this
with a
single downconversion placed in the receiver, using classical CATV tuner
technology
in front of the demodulation function in the receiver.
In a second embodiment of the invention, illustrated by Figure 8, much of the
same components as those of Figure 7 are implemented. However, as noted above,
in
the second embodiment DWDM transmitters are located in the node (865
collectively
in Figure 8), and are driven by the stacked RF signals. Accordingly, the
individual
wavelengths are transmitted back to the primary/secondary headend by ITU
transmitters (866 collectively).
At the primary/secondary headend, the optical signals are routed directly to
multiplexers 860 (it will be appreciated that since it is possible to have
different optical
levels, due to different node to OTN loss budgets, some level of signal
equalization
may be required). The output from the multiplexer 860 is sent to the master
headend in
the same manner as in the first embodiment. In addition, the primary/secondary
headend components of the second embodiment are assembled as those in the
first
embodiment as well.
One key advantage with the approach of the second embodiment is the reduced
amount of active equipment located within the primary/secondary headend. As
Figure
8 illustrates, it is no longer necessary to convert the optical signal back to
a RF signal.
Although this factor will improve performance, it places the transmitter in a
more
hostile environment in that temperature stability is one of the technical
issues
12
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
associated with not only the technology combination described herein, but with
DWDM itself.
Turning to a third embodiment of the invention, Figure 9 illustrates a
primary/secondary headend-based frequency stacking and DWDM system
implementation. As shown, both the frequency stacking system (shown as a 4 or
8
band system, the details of which have been discussed earlier herein) 900, and
the
DWDM lasers 910a-d/DWDM multiplexer 920, are all located within the
primary/secondary headend. The outputs of the fiber nodes are received at the
primary/secondary headend by dual receiver RPR/2 930.
Similarly, a slightly modified version of the third embodiment may also be
implemented. Referring back to the network architecture shown in Figure 2, at
the
secondary (or primary) hub, as in the third embodiment shown in Figure 9, the
reverse
path data may be aggregated to drive each DWDM laser transmitter using
Frequency
Stacking (FS) methods. The reverse path data transmission from each subscriber
is
typically one of the three basic multiple access schemes -- Code Division
Multiple
Access (CDMA), Frequency-Division-Multiple-Access (FDMA), Time-Division-
Multiple-Access (TDMA), or any combination of these schemes. Efficient use of
the
reverse path link, to ensure that the increased capacity is realized, uses any
combination
of CDMA, FDMA, TDMA to optimize the usage of the channel, together with the
combined DWDM/FS network architecture.
Accordingly, in each embodiment of the invention, the frequency stacking
system significantly increases the reverse path traffic capacity. This is
illustrated in
Figure 10, which shows a composite RF spectrum of four upconverted return path
frequency blocks (5-42 MHz). In this example, a reference pilot tone is
generated
above the payload multiplex in order to synthesize the four-band stack. The
pilot tone
is transmitted along with the upconverted signal and utilized in a block down-
converter
unit to synchronize the down-conversion, thus removing any frequency offset
errors.
The composite RF signal is then used to drive each of the DWDM reverse path
laser
transmitters. The DWDM laser transmitters can be either directly or externally
modulated DFB laser transmitters operating in 1550-nm wavelength band. As in
the
previous embodiments, multiplexer 920 optically multiplexes the signals from
DWDM
transmitters 91 Oa-d on a single fiber to be routed to the headend, where the
optical
13
CA 02370762 2001-10-18
WO 00/64087 PCT/US00/10358
signal may be amplified, and is optically demultiplexed to four different
optical
receivers. The composite output RF signal from each optical receiver is
transmitted to
a block down-converter unit, which extracts the four separate 5-42 MHz bands.
Again,
each of the high-speed data bands can be routed to return path application
receivers.
Figure l l illustrates the combined FSS/DWDM expansion process. As shown,
a single shared traditional 37 MHz segment provides 74 KHz per home (for 500
home
passed node). The implementation of frequency stacking (4 band) increases the
shared
segment to 148 Mhz thereby increasing the return bandwidth per home to 296KHz.
However, the implementation of both frequency stacking and DWDM increases the
return path bandwidth segment to 32 times the return path bandwidth segment,
or 1184
MHz, thereby increasing the return bandwidth per home to 2.368 MHz.
Accordingly, the architecture of the present invention provides an increased
capacity in the reverse path network and is well suited to implementation into
existing
systems that may have fiber limitations.
Although various embodiments are specifically illustrated and described
herein,
it will be appreciated that modifications and variations of the present
invention are
covered by the above teachings and within the purview of the appended claims
without
departing from the spirit and scope of the invention.
14
