Optical communications – Multiplex – Optical switching
Reexamination Certificate
2000-10-20
2004-08-24
Chan, Jason (Department: 2633)
Optical communications
Multiplex
Optical switching
C398S166000
Reexamination Certificate
active
06782200
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to the field of optical communications networks, and in particular to packet-based optical networks, and to the frame structure and node architecture used in such networks.
BACKGROUND OF THE INVENTION
The rapid increase in transmission capacity achieved by optical transmission systems has required the solution to numerous technical difficulties. The data capacity within optical networks is determined by the limits of transmission rates per channel and the number of available optical frequencies. The transmission capacities far exceed the capability of electronic processing of signals. However, electronic and opto-electronic elements are required for performing switching and routing functions per channel, and the conversion of many high-speed optical data to electrical signals with analysis to enable the switching and routing operations to be carried out is recognised as causing a limitation to data transfer rates.
There have been proposals which provide all-optical networks in which switching and routing take place in the optical domain. An alternative proposal is to provide packet based communication with the packet header optically encoded at a lower data rate than the data rate of the packet payload. This enables opto-electric conversion circuitry to be employed which has a lower detection bandwidth that that which would be required to carry out opto-electric conversion of the packet payload. Thus, low cost electronics can be used to enable the header to be read for routing purposes, and high speed conversion is required only when the payload data is to be read, at the destination node for the particular signal. Other advantages of packet-based optical transmission systems are well known. Generally, such systems allow much greater use of available bandwidth than circuit switched systems.
As optical pulses become more closely spaced and the channels of a WDM (wavelength division multiplex) system become more closely spaced along the frequency spectrum, the need for accurate control of optical pulse timing and shape is paramount. As the data transfer rates have increased, so has the ability to compensate for such effects as chromatic dispersion, either using appropriate fiber designs (such as dispersion managed fiber) or compensating elements (such as additional lengths of high positive dispersion fiber or Bragg gratings). These dispersion compensation measures enable the spectral spreading of pulses to be controlled and aligned accurately, even across different channels.
An optical network typically comprises a number of spaced nodes, which allow signals to be added to or dropped from the network. The nodes also perform routing functions, and have switching arrangements for this purpose. These may be optically transparent (photonic cross connects) or else the switching may take place in the electrical domain. These switching arrangements allow the signal at any input to be switched to any output. The complexity and cost of such arrangements is a function of the number of inputs and outputs, together with the number of times the data signal is converted from and to the optical domain.
Optical amplifiers are spaced between the nodes, for example every 80 km, and these compensate for optical signal attenuation. Dispersion compensation may also take place at the amplifier locations. In conventional networks, regeneration takes place at the nodes, which involves conversion from the optical domain to the electrical domain, with subsequent modulation of a new optical carrier. Proposed all-optical networks avoid the need for such electrical regeneration, but amplification and dispersion compensation is still required. More accurate dispersion compensation is required for higher bit rate systems.
If existing network architecture is to be upgraded to enable transmission at higher bit rates, this will typically require an increased number of amplifiers with shorter separation between amplifiers, in order to tolerate the higher bit rates. This upgrade requires significant expenditure on both transmission network and switching node hardware.
SUMMARY OF THE INVENTION
According to a first aspect of the invention, there is provided a WDM optical communications network comprising a plurality of nodes, wherein data to be transmitted between nodes is provided on a plurality of different wavelength channels structured as data sub-packets, wherein a plurality of the channels are grouped together to form a packet and wherein the format of one of the channels within the packet includes a packet header for providing routing or forwarding information for data on all of the channels of the group, the nodes of the network each comprising an optical switching device which selectively routes or forwards data packets at the inputs to the switching device to outputs of the switching device, wherein the data packets of the wavelength group of channels are provided to a single port of the switching device and are thereby routed together to an output port of the switching device in dependence on the contents of the packet header.
This network architecture groups packet channels together, with the channels of the group sharing a packet header. They can thus be provided to a single port of the cross connect, reducing usage of the ports of the cross connect. The invention enables an increase in the data transfer rate without altering the line equipment, as the individual channels still carry data at the same rate. This means the existing per band or per channel dispersion compensation elements and amplifier separation can be maintained.
The plurality of channels may comprise four consecutive DWDM channels.
Preferably, the optical switching device comprises a photonic cross connect which routes the packets without opto-electric conversion of the packet data. However, each node further comprises packet header reading circuitry having opto-electric conversion circuitry for converting the packet header from the optical to the electrical domain.
According to a second aspect of the invention, there is provided a packet structure for packetized optical data for transmission over a WDM optical network, wherein data to be transmitted between nodes is provided on a plurality of different wavelength channels structured as a data sub-packets, wherein a packet includes data for a plurality of the wavelength channels grouped together and a packet header, wherein the packet header provides common routing or forwarding information for the data on all of the wavelength channels of the group.
This wavelength group packet structure enables the network architecture of the invention to be implemented. The bit rate of the data in the packet header may be lower than the bit rate of data in the remainder of the packet. This enables simplified electronic circuitry to be used for reading the packet header, without the need to convert the user data to the electrical domain. The bit rate of the user data in the frame is greater than or equal to 10 Gb/s whereas the bit rate of data in the packet header is less than or equal to 10 Gb/s.
According to a third aspect of the invention, there is provided a node for a node for a WDM optical communications network, wherein data to be transmitted between nodes of the network is provided on a plurality of different wavelength channels structured as data sub-packets, the node comprising:
a partial demultiplexer for partially demultiplexing the channels at an input to the node into a plurality of groups of wavelength channels, each group of channels sharing header information;
header reading circuitry for reading the header information associated with the groups of channels;
an optical switching device for selectively routing data packets at the inputs of the switching device to the outputs of the switching device, where each group of channels forms a data packet which is routed together in dependence upon the header information associated with the group of wavelength channels; and
a multiplexer for multiplexing the switched groups of channels onto an out
Roorda Peter D
Warbrick Kevin
Barnes & Thornburg
Chan Jason
Leung Christina Y.
Nortel Networks Limited
LandOfFree
Packet-based optical communications networks does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Packet-based optical communications networks, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Packet-based optical communications networks will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3347312