# US 2021/0243125 A1 · System and method for avoiding congestion in a computer network

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2021/0243125 A1
- Document ID: US20210243125A1
- Status: Published application
- Kind: A1
- Application number: 17/122,539
- Filed: 2020-12-15
- Published: 2021-08-05
- Earliest claimed: 2020-01-31
- Assignees: Hewlett Packard Enterprise Development LP
- Applicants: Hewlett Packard Enterprise Development LP
- Inventors: Anil RAJ; Celestian Kaniampady SEBASTIAN; Jeswin AUGUSTINE
- Web page: https://patentgrep.com/patent/US20210243125A1

## CPC

- H04L 47/12: Avoiding congestion; Recovering from congestion
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06N 5/04: Inference or reasoning models
- G06N 20/00: Machine learning
- G06N 20/20: Ensemble learning
- H04L 45/026: {Details of "hello" or keep-alive messages}
- H04L 45/08: {Learning-based routing, e.g. using neural networks or artificial intelligence}
- H04L 47/127: by using congestion prediction
- H04L 47/28: in relation to timing considerations
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

The present invention relates a system and a method for avoiding congestion in a computer network. Information related to a flow of query data packets in the computer network and a capacity of the computer network is collected for processing. A pattern in the flow of query data packets is determined based on the collected information and a trained data model. The trained data model is used to determine occurrence of an impending connection data burst, in the computer network, which is capable of causing congestion in the computer network. The connection data burst comprises information related to links within routing devices present in the computer network. Upon determining occurrence of the impending connection data burst, parameters configured in the routing devices are modified to avoid the congestion of the computer network by the impending connection data burst.

## Claims

### Claim 1

A method comprising:
collecting information related to a flow of query data packets in a computer network and a capacity of the computer network;
determining a pattern in the flow of query data packets based on the collected information and a trained data model;
predicting occurrence of an impending connection data burst in the computer network using the trained data model, wherein the impending connection data burst is capable of causing congestion in the computer network, and wherein the impending connection data burst comprises information related to links within routing devices present in the computer network; and
modifying parameters configured in the routing devices to avoid the congestion of the computer network by the impending connection data burst, wherein the parameters comprise a waiting time for receiving the query data packets.

### Claim 2 (depends on 1)

The method as claimed in claim 1, wherein the trained data model is developed using a Machine Learning (ML) technique for processing the information related to the computer network.

### Claim 3 (depends on 2)

The method as claimed in claim 2, wherein the ML technique utilizes at least one of linear regression, Naïve Bayes, Principal Component Analysis (PCA), Decision Tree, Random Forest, and Gradient Boosting-Random Forest methods.

### Claim 4 (depends on 1)

The method as claimed in claim 1, wherein the query data packets are at least one of a Link State Advertisement (LSA) data packets and Hello data packets.

### Claim 5 (depends on 1)

The method as claimed in claim 1, wherein the information related to the links within the routing devices comprise a type of link, a cost of the link, and adjacencies with neighbouring routing devices.

### Claim 6 (depends on 1)

The method as claimed in claim 1, further comprising resetting the waiting time for receiving the query data packets to an original value ater the impending connection data burst is processed.

### Claim 7

A system comprising:
a processor; and
a memory connected to the processor, wherein the memory comprises programmed instructions which when executed by the processor, causes the processor to:
collect information related to a flow of query data packets in a computer network and a capacity of the computer network;
determine a pattern in the flow of query data packets based on the collected information and a trained data model;
predict occurrence of an impending connection data burst in the computer network using the trained data model, wherein the impending connection data burst is capable of causing congestion in the computer network, and wherein the impending connection data burst comprises information related to links within routing devices present in the computer network; and
modify parameters configured in the routing devices to avoid the congestion of the computer network by the impending connection data burst, wherein the parameters comprise a waiting time for receiving the query data packets.

### Claim 8 (depends on 7)

The system as claimed in claim 7, wherein the trained data model is developed using a Machine Learning (ML) technique for processing the information related to the computer network.

### Claim 9 (depends on 8)

The system as claimed in claim 8, wherein the ML technique utilizes at least one of linear regression, Nave Bayes, Principal Component Analysis (PCA), Decision Tree, Random Forest, and Gradient Boosting-Random Forest methods.

### Claim 10 (depends on 7)

The system as claimed in claim 7, wherein the query data packets are at least one of a Link State Advertisement (LSA) data packets and Hello data packets.

### Claim 11 (depends on 7)

The system as claimed in claim 7, wherein the information related to the links within the routing devices comprise a type of link, a cost of the link, and adjacencies with neighbouring routing devices.

### Claim 12 (depends on 7)

The system as claimed in claim 7, further comprising resetting the waiting time for receiving the query data packets to an original value after the impending connection data burst is processed.

### Claim 13

A non-transitory machine readable storage medium having stored thereon machine readable instructions to cause a computer processor to:
collect information related to a flow of query data packets in a computer network and a capacity of the computer network;
determine a pattern in the flow of query data packets based on the collected information and a trained data model;
predict occurrence of an impending connection data burst in the computer network using the trained data model, wherein the impending connection data burst is capable of causing congestion in the computer network, and wherein the impending connection data burst comprises information related to links within routing devices present in the computer network; and
modify parameters configured in the routing devices to avoid the congestion of the computer network by the impending connection data burst, wherein the parameters comprise a waiting time for receiving the query data packets.
