Electrical computers and digital processing systems: multicomput – Computer network managing
Reexamination Certificate
1998-09-04
2002-04-09
Harrell, Robert B. (Department: 2152)
Electrical computers and digital processing systems: multicomput
Computer network managing
Reexamination Certificate
active
06370572
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Technical Field of the Invention
The present invention relates in general to the communications field and, in particular, to a system and method for improving the performance management of a distributed communications network.
2. Description of Related Art
The tasks of managing and controlling the performance of distributed communications networks (e.g., distributed data networks or distributed telecommunications networks) are becoming increasingly complex due to a number of crucial factors, such as, for example, the increased complexity, dynamism and diversity of the network technologies, the spread of advanced services with very distinct requirements (e.g., live video, file transfers, etc.), and the heightened expectations of the users being served. Other crucial factors that impact network complexity are the progressive deregulation of the telecommunications industry, and the highly competitive market that has emerged as a result.
In order to survive in such an environment, a distributed communications network operator must manage the network so that its utilization is maximized (i.e., ensure a maximum return on the investment), while ensuring that all offered services perform within expected bounds. In order to perform such tasks, the operator's personnel need certain support tools that help them to manage the tasks with their complexities. In particular, certain distributed, dynamically changing networks, such as, for example, the next generation Internet and so-called third generation mobile communication networks will require a level of operational support that is not provided by today's support systems.
Operation and Support Systems (OSS) typically function to perform routine support tasks in data communications and telecommunications systems, such as, for example, traffic measurements, network supervision and performance management, analyses, fault diagnoses, administrative tasks, etc. The current approach used for network performance management in the OSS industry typically involves a number of applications residing on a software platform. The software platform usually supports separate applications for monitoring network performance information, managing alarm conditions, and handling of common functions in order to initiate management operations for network resources. Normally, these applications are not integrated to a great extent, other than that they share the same platform facilities. Consequently, it is the operator who has to correlate the performance and alarm information, and where necessary, decide what actions are appropriate to take with regard to improving network performance.
As such, most of the support systems involved are centralized in a single, monolithic management center, or in some cases, distributed or spread across a relatively small number of geographically distinct management centers. In some of the distributed system cases, the main reason for the distribution is the distributed nature of the responsibilities in the corporate organizations involved.
Currently, in a typical telecommunication system, the network element of the system gathers statistics about the traffic it is handling over a five or fifteen minute interval. The network element then makes this information available to the system as an output file, or stores it locally for later retrieval. As such, two of the original motives for structuring the telecommunication system performance measurement activities in this way were to minimize the sheer volume of information generated, and reduce the network element's processor load. Typically, the performance information is retrieved by a network element's management system, and stored in a database from which performance reports can be generated, either periodically or on demand.
A relatively small subset of management/control functions are automated in today's telecommunications networks. These functions are almost exclusively implemented in the network elements, and their scope is typically confined to the specific network element involved. An example of such a management/control function is “overload control,” which attempts to regulate the load on a network element so that the element's performance is maintained within acceptable bounds.
Nevertheless, there are a number of problems that exist with the current state of the art performance management solutions. However, as described below, the problems particularly relevant to the present invention are related to management/control complexity, timing and centralization. For example, telecommunication networks are becoming increasingly complex to manage in terms of their scale, the diversity of the networks and services they provide, and the resulting voluminous amount of information that must be handled by the management system. In order to address these complexities, certain semi-automated and automated management solutions will be needed to support a network operator's staff. However, such support capabilities actually do not exist in the management solutions provided today.
Specifically, today's performance management systems collect network performance measurement information at fixed intervals (granularities) of 5 minutes, 15 minutes, or 1 hour. However, implementation at such fixed intervals effectively introduces an inherent latency or delay in the availability of the performance information. Consequently, these delays effectively limit the ability of network managers to analyze the performance of their networks. Clearly, in operating dynamic telecommunication networks such as cellular networks, Internets, and broadband multi-media networks, these delays in identifying and resolving network performance problems are unacceptable. Furthermore, as the network management/control systems become increasingly automated, such delays in the delivery of performance measurement information, and the coarse granularity of the measurements themselves, will become increasingly unacceptable. Instead, the measurement and control intervals used should be dictated by the timing requirements of the problem domain, rather than by the solutions the network elements provide today.
Today's telecommunication network management systems are deployed in a relatively small number of locations in the network. In other words, the management functions are centralized in a small number of network nodes. This approach works to a certain extent in today's large telecommunication networks, because the network functions are also centralized in very large and costly network elements. As mentioned earlier, the control algorithms (e.g., overload control) that exist today are limited in scope to a single network element. However, in the distributed processing networks of the future, this single network element approach will not be acceptable because the distribution of service logic across, and interdependencies between, the distributed network elements will require control solutions that can span the distributed nodes.
SUMMARY OF THE INVENTION
In accordance with the preferred embodiment of the present invention, a distributed communications network management and control system includes a real time control system, which is composed of a plurality of control agents. The control agents receive real time performance information associated with the distributed communications network from a plurality of performance agents, analyze the real time performance information, and in due time, output control commands that control the performance of the distributed communications network based on the real time performance data received.
An important technical advantage of the present invention is that it provides a system that supports the abstraction of managed network information to support views that are appropriate for solving distinct control problems.
Another important technical advantage of the present invention is that it provides a system that is both flexible and scalable, with the possibility of configuring the system to suit the needs o
Lindskog Peter
Newcombe Adrian
Harrell Robert B.
Jenkens & Gilchrist P.C.
Telefonaktiebolaget L M Ericsson (publ)
LandOfFree
Performance management and control system for a distributed... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Performance management and control system for a distributed..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Performance management and control system for a distributed... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2929449