Electrical computers and digital processing systems: multicomput – Computer network managing
Reexamination Certificate
2000-09-26
2004-05-04
Burgess, Glenton B. (Department: 2153)
Electrical computers and digital processing systems: multicomput
Computer network managing
C709S224000, C709S225000, C714S057000
Reexamination Certificate
active
06732169
ABSTRACT:
The present invention relates to a network transmission circuit control method and system. The network in question is essentially a regional telecommunications network,. i.e. a network able to administer a large number of users, for example around one million users.
BACKGROUND OF THE INVENTION
At national level, telephone operators organize calls within their general network by means of networks which typically include a thousand routing and drop-and-insert nodes, also referred to as network elements (NE). The regions to which the networks relate are either large provinces or large towns, the number of users to be connected being no more than an inaccurate indication. The routing nodes of these networks are generally interconnected by STM16 links, i.e. links capable of transmitting data at a bit rate of 2.5 Gbit/s. The links can also be STM1 links (an STM16 link is equivalent in terms of bit rate to sixteen STM1 links) which offer bit rates of 155 Mbit/s, 3×34 Mbit/s, 3×45 Mbit/s, 21×6 Mbit/s or 63×2 Mbit/s, depending on how they are configured. These bit rates are used to construct architectures matched as closely as possible to requirements. The plant installed at each node and the plant configurations depend on local requirements in the vicinity of the node. SDH networks are networks of the above kind. Some network nodes have their own special features.
Some nodes are connected to inter-regional optical fiber links, for example transoceanic links, in particular links under the Atlantic Ocean. In this case the bit rates encountered are of the order of 40 Gbit/s for each optical fiber. An optical fiber link contains several optical fibers capable of bit rates of this magnitude.
Other nodes also have functions complementary to the routing and drop-and-insert functions which are specific to the other nodes. These other nodes are gateway nodes, which are also referred to as gateway network elements (GNE), and connect a network management unit to each network node.
The network management unit, which is referred to hereinafter as the network management system, includes signaling links for organizing the interconnection and working of the routing and drop-and-insert nodes of the network. The information on these signaling links can also be routed via normal channels of the network.
Because an SDH network has a meshed structure, any two points of the network can be interconnected via multiple paths. If a network link fails, the traffic carried by that link can be diverted to other links between the same departure and arrival points. Transmission is modified at the request of clients who are either operators, i.e. physical persons operating the controls of the network management system (keyboard, mouse, or other devices), or automatic processes initiated on demand by the network management system when it receives from a routing node a reorganization request concerning that node. For example, for the bit rates indicated above, the plant at a node can be modified to change it from one architecture and a given bit rate to another architecture and another bit rate, for example from 3×45 Mbit/s to 21×6 Mbit/s. These configuration changing tasks are never urgent and can be deferred for a few seconds or even a few minutes. They are the responsibility of the network management system which, to prevent overloads, assigns priorities to the tasks to be undertaken.
For example, input/output management must be processed with a higher priority than configuration management or network reorganization. A third type of action has an even higher priority, namely actions programmed at fixed times, for example to verify the status of connections to the network nodes.
To manage priority levels, a central system of the network management system distributes the tasks to be effected by the network management system between a number of stacks, namely a highest priority synchronization (timer) stack, an input/output management stack and a deferred jobs stack. The network management system processes jobs which have accumulated in a stack of relatively lower priority only if all jobs which have accumulated in stacks of relatively higher priority have been processed. Accordingly, each time the network management system receives a request, an input/output request or a synchronization instruction, it stores it in the corresponding stack and processes it according to its priority.
Some events, namely alarms which can be of many different kinds, receive special processing within this hierarchy. Obviously they do not have an a priori priority level. Nevertheless, because they constitute events affecting the continued existence of the network, and potential total breakdown of the network, alarms must be processed with the maximum priority, essentially so as at least to give them a severity level for assessing the scale and the urgency of the reaction to the fault that the alarms represent. Accordingly, the central system of the network management system processes all alarms as soon as they arrive, giving them priority over other events.
It is apparent that this type of operation leads in some cases to blockage of the network management system. If a high bit rate link deteriorates, for example a transoceanic optical fiber link of the kind referred to above, the number of “transmission impossible” alarms allocated to transmission channels using that optical fiber becomes very large. For a given network management system architecture, an alarm processing rate of the order of 20 alarms per second has been adopted. Because in practice the network management system processes not only alarms but also other actions, the usable alarm processing rate is of the order of five alarms per second. If a very high bit rate link fails a thousand or even several thousand alarms may be sent.
With a number of alarms of this magnitude, the network management system is immobilized for about twenty minutes while it processes all the alarms. Operations requested of the network management system cannot be processed before the end of this time. This can block the network management system and suspend the network supervision service.
A solution entailing oversizing the network management system plant is not economically viable because massive alarm situations do not arise frequently, and because it would entail providing a plethora of processing means whose own reliability could become a problem in itself, and might even run the risk of not leading to a reasonable solution to the blocking problem.
OBJECTS AND SUMMARY OF THE INVENTION
The invention proposes a simpler solution to the problem. It has been realized that, when a fault occurs, the reaction to the fault could not be rapid in any event. An optical fiber cannot be repaired by remote control. The invention therefore divides the processing of initially monolithic alarms (“monolithic” in the sense that several dozen alarms can be contained in a single message incoming into the system) which have high priority into a succession of shorter processes which take up less time (which deal with five alarms at a time, for example), and allocates lower priority to these processes. Accordingly, and keeping a sense of proportion, assessing the severity of an alarm is in itself regarded as a non-urgent task. Consequently, according to the invention, if one or more alarms occur, a central system of the network management system executes (or causes to be executed) a very fast task which can itself be one of the highest priority tasks and which breaks alarms in the incoming stream of alarms down into lower priority tasks, each responsible for processing a smaller number of alarms, for example five alarms. The processing entails forwarding the alarms in a more complete format. This prevents the physical persons operating the system losing control of the network management system.
The invention therefore provides a method of managing transmission circuits of a network, in which method:
a network management system receives configuration and/or connection requests and, from
Heritier Fabien
Lapraye Bertrand
Thome Franck
Alcatel
Barqadle Yasin
Burgess Glenton B.
Sughrue & Mion, PLLC
LandOfFree
Network transmission circuit control method and system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Network transmission circuit control method and system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Network transmission circuit control method and system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3200053