Multiplex communications – Channel assignment techniques – Using time slots
Reexamination Certificate
1998-10-29
2004-05-04
Nguyen, Steven H. D (Department: 2665)
Multiplex communications
Channel assignment techniques
Using time slots
C370S249000, C370S254000, C709S241000, C709S235000
Reexamination Certificate
active
06731648
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a communications network and to nodes of such a network, and, in one aspect, in particular to an optical communications network using packet routing to transmit data in the optical domain at very high bit rates.
BACKGROUND OF THE INVENTION
The increasing processing power and storage capacity of computers has lead to the development of, e.g., multimedia applications that generate high bandwidth data. This in turn has lead to a need for broadband networks capable of handling the traffic generated by such applications. Desirably, such networks should be capable of handling bursty high-bandwidth data traffic virtually instantaneously on demand.
Photonic networks have been developed to meet the need for broadband data communications. However, with current technologies there is a mismatch between the capabilities of optical networks and the types of services required by users. Typically, existing optical networks use wavelength-routed circuit connections which are ill-adapted for bursty data traffic and connectionless applications. Currently, this mismatch is handled by overlaying the optical network with electronic service layers using, for example, IP (internet protocol), ATM (asynchronous transport mode), frame relay or SDH (synchronous digital hierarchy) protocols. Such approaches become increasingly inadequate as computer processing power and the associated bandwidth requirements increase still further, and as new applications are developed using, for example, intelligent agents which search and retrieve remote information, process the information and present the information to the user through rich and intelligent user interfaces.
SUMMARY OF THE INVENTION
According to a first aspect of the present invention, there is provided a method of operating an optical communications system comprising a plurality of routing nodes, each routing node being configurable to direct a received optical packet onwards via a selected one of two or more output paths, and an optical network interconnecting the plurality of routing nodes, the method comprising
a) outputting an optical packet from an originating routing node onto the optical network;
b) configuring the optical network and the routing nodes to provide a looped transmission path between the originating routing node and a destination routing node;
c) receiving the optical packet at the destination routing node; and
d) transmitting a return signal to the originating routing node in a time slot on the looped transmission path which was occupied by the optical packet output in step (a).
This aspect of the present invention provides a method of operating a communications network which supports packet transmission while offering good reliability and very low latency. The invention also makes efficient use of the broad bandwidth available on optical networks, and also avoids the technological problems in this context of prior art methods which require optical buffering and complex bit-level processing. Reliable communication generally requires a handshake between originating and destination nodes. The invention reduces the time taken for this handshake to essentially the round trip time between the originating and destination nodes. This is achieved with a looped signal path. The same time slot on the looped signal path is used for both the outgoing transmission of a packet and the return transmission of the acknowledgement. As well as minimising latency, this method has the further advantage that the originating node can predict precisely when the acknowledgement signal is expected, and so can glean useful information from the absence of the acknowledgement signal at the expected time. The absence of an acknowledgement signal at the expected time can trigger a transmission failure event resulting, for example, in the retransmission of the relevant packet or string of packets.
Preferably the return signal comprises one of a plurality of packets received on the looped signal path from the originating node.
In preferred implementations of the invention, the signal transmitted to the destination node comprise a string of packets and the destination node strips off the payload packets and returns on the transmission path, e.g., a header packet. The returned packet may be modified, for example by overwriting one or more bits of an acknowledgement flag. Alternatively, the destination may generate a new packet for transmission back to the originating node. The return signal is not necessarily a simple acknowlegement signal but may, for example, comprise data to be transmitted to the source node in response to a polling signal. In this case the method may include:
transmitting a polling signal from the source node to the destination node:
scheduling a process at the source node for execution at a time after the transmission of the polling signal, which time is dependent on the return trip time between the source node and the destination node;
transmitting data from the destination node to the source node in the time slot occupied by the said polling signal; and
subsequently executing the said process using the said data.
Preferably the network has a mesh topology. It may be fully meshed or partially meshed or may have a hybrid topology. Preferably the network comprises a multiplicity of nodes and links, and the nodes and links are configured as a multiplicity of directed trails, each directed trail linking some only of the multiplicity of nodes and the directed trails in combination spanning every node of the network and the looped signal path comprises a closed directed trail which includes both the originating node and the destination node.
Although the present invention in its broader aspects is by no means limited to use with any one network topology nor any one routing scheme, it is particularly advantageous to combine the method of the invention with the directed trail routing method which is described and claimed in the present applicant's copending international application, also entitled “Communications Network”, WO 98/09403 Agent's reference A25265/WO. The network is then configured as a number of directed trails, and packets are routed by the originating node selecting one of the trails which spans the originating and destination nodes. No processing other than simple address recognition is then required at the intermediate nodes. The directed trail method offers low latency, good scalability and very low processing overheads. When combined with the method of the present invention, it makes it possible for the entire process of routing, receiving and acknowledging to be carried out at very high speeds.
According to a second aspect of the present invention, there is provided a method of operating a node in a communications network comprising
(a) receiving a packet from a looped signal path; and
(b) outputting a return signal onto the looped signal path, in the time slot originally occupied by the said packet.
According to a third aspect of the present invention, there is provided a node suitable for connection in a communications network and comprising:
a) an output arranged to output a packet onto a looped signal path for transmission to another node;
b) an input arranged to receive any signal returned by the other node on the looped signal path;
c) a node controller arranged to monitor signals returned to the said input.
According to a fourth aspect of the present invention, there is provided a method of operating a node in a communications network, the method comprising:
transmitting a polling signal from the source node to the destination node:
scheduling a process at the source node for execution at a time after the transmission of the polling signal, which time is dependent on the return trip time between the source node and the destination node;
transmitting data from the destination node to the source node in the time slot occupied by the said polling signal; and
subsequently executing the said process using the said data.
In this fourth aspect the network may be an
British Telecommunications PLC
Nguyen Steven H. D
Phan M.
LandOfFree
Method and apparatus for transmitting a return signal in a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for transmitting a return signal in a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for transmitting a return signal in a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3255711