SD-WAN Link Selection from Network Slice Information

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing SD-WAN technologies face challenges in selecting optimal links for different types of traffic, leading to issues such as decreased throughput and increased packet latency due to the limitations of existing algorithms like deep packet inspection.

Innovation Solution

An SD-WAN controller uses network slice information to determine link selection rules by performing shallow packet inspection and associating IP addresses with transmission characteristics, enabling more efficient link selection based on specific traffic requirements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If deep packet inspection algorithms are used to select WAN links for different traffic flows, then link selection accuracy is improved, but network performance deteriorates due to decreased throughput and increased packet latency

Engineering Contradiction:
Improvelink selection accuracyVSAvoidnetwork throughput
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent extracts only the necessary packet header information (source IP address, destination IP address, protocol type) required for link selection, rather than performing deep inspection of the entire packet. This extraction approach maintains link selection accuracy while significantly reducing processing overhead and improving network throughput

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system pre-establishes link selection rules that map specific IP address combinations and protocol types to optimal WAN links before traffic arrives. This preliminary configuration allows the router to make rapid link selection decisions based on pre-analyzed traffic patterns, avoiding real-time deep packet inspection and reducing packet latency

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If deep packet inspection is performed to identify traffic types for link selection, then traffic classification accuracy is improved, but packet processing time increases leading to higher latency

Engineering Contradiction:
Improvetraffic classification accuracyVSAvoidpacket latency
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent extracts only the essential packet header fields (source IP address, destination IP address, protocol type) needed for traffic classification and link selection. By extracting only these critical elements rather than performing deep inspection of packet payloads, the system maintains accurate traffic classification while minimizing packet processing time and latency

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system performs preliminary traffic classification by establishing rules that map IP address combinations and protocol types to traffic categories and optimal links before packets arrive. This advance preparation enables rapid classification decisions without requiring time-consuming deep packet inspection, thereby reducing packet latency

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If multiple link selection rules are maintained for different traffic flows, then link selection precision is improved, but system complexity increases

Engineering Contradiction:
Improvelink selection precisionVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the link selection process into distinct, manageable components: packet header extraction, rule matching based on source/destination IP addresses and protocol type, and link selection. This segmentation allows multiple precision rules to be maintained without overwhelming system complexity, as each component handles a specific aspect of the selection process independently

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP4302468B1Methods, systems, and computer readable media for selecting a software defined wide area network (sd-wan) link using network slice information
Publication Date: 2025.10.08 ORACLE INT CORP
  • EP4302468B1 patent drawingFigure 1
  • EP4302468B1 patent drawingFigure 2
  • EP4302468B1 patent drawingFigure 3

AI summary

Methods, systems, and computer readable media for selecting a software defined wide area network (SD-WAN) link using network slice information are disclosed. One method occurs at an SD-WAN controller implemented using at least one processor, the method comprises: receiving a packet containing packet header information including at least one internet protocol (IP) address; determining, using the packet header information, a link selection rule for selecting an SD-WAN link for transmission, wherein the link selection rule uses transmission characteristics associated with the packet, wherein the transmission characteristics are discerned using network slice information associated with the packet; selecting, using the link selection rule, the SD-WAN link; and transmitting the packet via the SD-WAN link.