Multiplex communications – Data flow congestion prevention or control – Control of data admission to the network
Reexamination Certificate
1997-06-16
2002-04-23
Nguyen, Chau (Department: 2663)
Multiplex communications
Data flow congestion prevention or control
Control of data admission to the network
C370S395200
Reexamination Certificate
active
06377545
ABSTRACT:
TECHNICAL FIELD OF THE INVENTION
The present invention relates an access control mechanism, for a telecommunications device, for example an ATM switch.
DESCRIPTION OF RELATED ART
It is widely expected that ATM networks will enable the satisfaction of performance requirements of a whole variety of services, such as voice, video, and data on a single infrastructure. To achieve this, ATM networks must provide traffic control management.
A problem which arises in ATM switches is due to the random nature of ATM traffic. The switch control elements must be able to cope with variations in traffic.
The performance of an ATM network can be defined in terms of the Quality of Services (QoS) requirements which must be delivered to the network subscriber. These performance measures include: the network throughput, latency, jitter or delay variation, and the amount of cell loss that can be tolerated.
A network provider may agree to handle a certain number of cells (units of traffic) from a particular subscriber. It is then necessary for the network provider to have an access control system, which verifies that the subscriber is not supplying excess traffic to the network. In the limit, this excess traffic might not be allowed to access the network, depending on demands from other users.
SUMMARY OF THE INVENTION
The present invention provides a method for open-loop adaptive control of access to a broadband packet data type telecommunications network. Based on expected future data arrival patterns, data which violate agreed quality of service parameters may be denied access to the network.
The present invention relates to the use of neural networks, specifically pRAM (probabilistic RAM), networks in ATM switch design, as part of the access control system. Since the pRAM learns very fast, and has been shown to posses excellent generalisation properties in noisy environments, it is very suitable to cope with the stochastic nature of broadband ATM traffic. A pyramidal probabilistic RAM, configured in a teacher forcing mode, provides the required control. The pRAM may be trained off-line, before being integrated as the control element of an ATM node, or on-line.
Specifically, the neural network is trained to be able to predict incoming traffic flows to the switch. If incoming data is such that it is probable, based on the expected future traffic, that such future traffic will be unable to access the switch, then the incoming data may be denied access to the switch.
REFERENCES:
patent: 5263121 (1993-11-01), Melsa et al.
patent: 5287347 (1994-02-01), Spanke
patent: 5357507 (1994-10-01), Hughes et al.
patent: 5497375 (1996-03-01), Hluchyj et al.
patent: 5511066 (1996-04-01), Witters et al.
patent: 5564115 (1996-10-01), Clarkson
patent: 5687292 (1997-11-01), Boda et al.
patent: 5691975 (1997-11-01), Hamada et al.
patent: 5796735 (1998-08-01), Miller et al.
patent: 5825750 (1998-10-01), Thompson
patent: 5828653 (1998-10-01), Goss
patent: 94/24637 (1994-10-01), None
patent: 95/34979 (1995-12-01), None
Guan et al, The application of noisy reward/penalty learning to pyramidal pRAM structure, IEEE, 1992.*
Lee et al, A Neutral Network Based ATM Call Admission Controller for Multiple Service Classes with Different Qos, IEEE, May 1996.*
Sarajedini et al, Quality of Service Prediction using Neural Network, IEEE, Aug. 1990.*
J, Austin, A review of RAM based neural networks, IEEE, Sep. 1994.*
Onyiagha G. et al., “Probabilistic-RAM Neural Networks In ATM Traffic Shaping and Policing”, Proc. Int. Conf. on Engineering Applications of Neural Networks, Jun. 17-19, 1996, pp. 229-232.
Godfrey Onyiagha et al., “Adaptive Access Control of ATM Traffic Using Neural Networks”, IEEE Document No. 0-7803-3336-5/96, Nov. 16, 1996, pp. 201-205.
Onyiagha G. et al., “Adaptive Access Control of ATM Traffic Using Neural Networks”, A Collaborative Research Project Between Ericsson Components Ltd. and King's College London Under the PRI Scheme, Nov. 16, 1996, pp. 1-18.
United Kingdom Patent Office Search Report.
Burns Doane Swecker & Mathis L.L.P.
Lee Andy
Nguyen Chau
Telefonaktiebolaget LM Ericsson (publ)
LandOfFree
Open loop adaptive access control of ATM networks using 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 Open loop adaptive access control of ATM networks using a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Open loop adaptive access control of ATM networks using a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2909761