Automatic detection and prevention of network overload conditions using SDN

• US 10,432,533 B2
• Filed: 04/25/2017
• Issued: 10/01/2019
• Est. Priority Date: 10/21/2014
• Status: Expired due to Fees

• Current Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Original Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Inventors: Park, Sun-Uk, Huynh, Minh, Mangla, Ashima
• Primary Examiner(s): Georgandellis, Andrew C
• Application Number:
  US15/496,557
• Publication Number:
  US 20170230297A1
• Time in Patent Office: 889 Days
• CPC Class Codes: H04L 41/0681 Configuration of triggering conditions
• CPC Class Codes: H04L 41/0816 the condition being an adaptation, e.g. in response to network events
• CPC Class Codes: H04L 41/0896 Bandwidth or capacity management, i.e. automatically increasing or decreasing capacities flow or congestion control using dynamic resource allocation, e.g. in-call renegotiation, H04L47/76
• CPC Class Codes: H04L 41/0897 by horizontal or vertical scaling of resources, or by migrating entities, e.g. virtual resources or entities
• CPC Class Codes: H04L 41/147 for predicting network behaviour
• CPC Class Codes: H04L 41/40 using virtualisation of network functions or resources, e.g. SDN or NFV entities
• CPC Class Codes: H04L 43/16 Threshold monitoring
• CPC Class Codes: H04L 43/20 the monitoring system or the monitored elements being virtualised, abstracted or software-defined entities, e.g. SDN or NFV
• CPC Class Codes: H04L 47/127 by using congestion prediction

Abstract

A controller node in a network implementing software defined networking (SDN) proactively detects network overload conditions in a target node and subsequently prevents the conditions from developing by launching a new virtual network node. The detection and prevention are based on the determination of a probability of failure of the target node as a weighted function of information from the target node and probabilities of failure of related nodes.

Claims

1. A method for predicting a network overload condition at a target network node in a communications network, the target network node being a policy and charging rules function node, the method comprising:

collecting key performance indicators from a plurality of source network nodes, the source network nodes being edge nodes;

based on the key performance indicators collected from the plurality of source network nodes, computing probabilities of failure of each of the source network nodes;

collecting key performance indicators from a plurality of intermediate network nodes, the intermediate network nodes including at least one of a gateway node and a path computation element function node, each intermediate network node receiving information from at least one of the source network nodes;

based on the key performance indicators collected from the intermediate network nodes, and further based on the probabilities of failure of the source network nodes from which information is received, computing probabilities of failure of each of the intermediate network nodes;

collecting key performance indicators from the target network node, the target network node receiving information from the intermediate network nodes;

by a controller node of the communications network, predicting the network overload condition when a probability of failure P(F) of the target network node exceeds a threshold, the probability of failure of the target network node being a function of the key performance indicators I_F of the target network node and further being a function of the probabilities of failure P(D), P(E) of the intermediate network nodes, weighted by a closeness of relationships of the intermediate network nodes to the target network node;

wherein the probability of failure of the target network node is described by

P(F)=ƒ

(I_F,P(D),P(E))

 

wherein the function ƒ

is varied according to a manufacturer of the target node.

2. The method of claim 1, wherein the probabilities of failure of each of the source network nodes are computed without using information from other nodes.

3. The method of claim 1, wherein the closeness of relationships to the target network node includes an aggregation level and topography of the intermediate network node.

4. The method of claim 1, further comprising:

by the controller node of the communications network, backtracking from the network overload condition of the target network node to determine a node that is a probable source of the network overload condition;

based on predicting the network overload condition, spawning a new instance of the node that is the probable source of the network overload condition; and

notifying a support node of the communication network that the new instance is accessible to the communications network.

5. The method of claim 4 wherein the support node is a domain name system server.

6. The method of claim 1 wherein the controller node is a Diameter routing agent.

7. A network controller node of a communications network, comprising:

at least one processor;

at least one network interface; and

at least one computer readable storage device having stored thereon computer readable instructions that, when executed by the at least one processor, cause the at least one processor to perform operations for predicting a network overload condition at a target network node in the communications network, the target network node being a policy and charging rules function node, the operations comprising;

collecting key performance indicators from a plurality of source network nodes, the source network nodes being edge nodes;

based on the key performance indicators collected from the plurality of source network nodes, computing probabilities of failure of each of the source network nodes;

collecting key performance indicators from a plurality of intermediate network nodes, the intermediate network nodes including at least one of a gateway node and a path computation element function node, each intermediate network node receiving information from at least one of the source network nodes;

based on the key performance indicators collected from the intermediate network nodes, and further based on the probabilities of failure of the source network nodes from which information is received, computing probabilities of failure of each of the intermediate network nodes;

collecting key performance indicators from the target network node, the target network node receiving information from the intermediate network nodes;

predicting the network overload condition when a probability of failure P(F) of the target network node exceeds a threshold, the probability of failure of the target network node being a function of the key performance indicators I_F of the target network node and further being a function of the probabilities of failure P(D), P(E) of the intermediate network nodes, weighted by a closeness of relationships of the intermediate network nodes to the target network node;

wherein the probability of failure of the target network node is described by

P(F)=ƒ

(I_F,P(D),P(E))

wherein the function ƒ

is varied according to a manufacturer of the target node.

8. The network controller node of claim 7, wherein the probabilities of failure of each of the source network nodes are computed without using information from other nodes.

9. The network controller node of claim 7, wherein the closeness of relationships to the target network node includes an aggregation level and topography of the intermediate network node.

10. The network controller node of claim 7, wherein the operations further comprise:

backtracking from the network overload condition of the target network node to determine a node that is a probable source of the network overload condition;

based on predicting the network overload condition, spawning a new instance of the node that is the probable source of the network overload condition; and

notifying a support node of the communication network that the new instance is accessible to the communications network.

11. The network controller node of claim 10, wherein the support node is a domain name system server.

12. The network controller node of claim 7, wherein the controller node is a Diameter routing agent.

13. A non-transitory computer-readable medium having stored thereon computer readable instructions for predicting a network overload condition at a target network node in a communications network, the target network node being a policy and charging rules function node, wherein execution of the computer readable instructions by a processor causes the processor to perform operations comprising:

collecting key performance indicators from a plurality of source network nodes, the source network nodes being edge nodes;

based on the key performance indicators collected from the plurality of source network nodes, computing probabilities of failure of each of the source network nodes;

collecting key performance indicators from a plurality of intermediate network nodes, the intermediate network nodes including at least one of a gateway node and a path computation element function node, each intermediate network node receiving information from at least one of the source network nodes;

based on the key performance indicators collected from the intermediate network nodes, and further based on the probabilities of failure of the source network nodes from which information is received, computing probabilities of failure of each of the intermediate network nodes;

collecting key performance indicators from the target network node, the target network node receiving information from the intermediate network nodes;

predicting the network overload condition when a probability of failure P(F) of the target network node exceeds a threshold, the probability of failure of the target network node being a function of the key performance indicators I_F of the target network node and further being a function of the probabilities of failure P(D), P(E) of the intermediate network nodes, weighted by a closeness of relationships of the intermediate network nodes to the target network node;

wherein the probability of failure of the target network node is described by

P(F)=ƒ

(I_F,P(D),P(E))

wherein the function ƒ

is varied according to a manufacturer of the target node.

14. The non-transitory computer-readable medium of claim 13, wherein the probabilities of failure of each of the source network nodes are computed without using information from other nodes.

15. The non-transitory computer-readable medium of claim 13, wherein the closeness of relationships to the target network node includes an aggregation level and topography of the intermediate network node.

16. The non-transitory computer-readable medium of claim 13, further comprising:

backtracking from the network overload condition of the target network node to determine a node that is a probable source of the network overload condition;

based on predicting the network overload condition, spawning a new instance of the node that is the probable source of the network overload condition; and

notifying a support node of the communication network that the new instance is accessible to the communications network.

17. The non-transitory computer-readable medium of claim 16, wherein the support node is a domain name system server.