Semantic Background Knowledge Graph Alignment for Reliable Communication

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing semantic communication systems lack mechanisms for updating background knowledge, which is essential for proper interpretation and alignment of semantic representation vectors, leading to inefficiencies in semantic communication.

Innovation Solution

A method and apparatus for constructing a relation between background knowledge and semantic representation vectors, ensuring alignment and uniformity on a unit hypersphere, involving synchronization signals, system information exchange, and configuration of a background knowledge graph to facilitate semantic communication.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If background knowledge is not updated in semantic communication systems, then the system structure remains simple and operation is easy, but semantic interpretation accuracy deteriorates and communication reliability decreases

Engineering Contradiction:
Improvesemantic interpretation accuracyVSAvoidsystem structure complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent performs preliminary actions by pre-configuring background knowledge graphs and semantic representation vectors before communication occurs. The system pre-establishes the structural framework for knowledge representation and pre-aligns semantic vectors on hyperspheres, so that when communication happens, the interpretation can proceed accurately without requiring complex real-time updates of the fundamental structure.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent segments the background knowledge into structured knowledge graphs with distinct nodes and edges, and separates semantic representation into independent vectors that can be individually aligned and updated. This segmentation allows the system to maintain simplicity at the structural level while enabling targeted updates of specific knowledge elements without complicating the overall system architecture.

Inventive Principle:
Principle #1Segmentation

2Reliability

If background knowledge is updated through relation configuration between background knowledge and semantic representation vectors, then semantic communication reliability improves, but the operation complexity and protocol requirements increase

Engineering Contradiction:
Improvesemantic communication reliabilityVSAvoidoperation complexity
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The patent changes parameters by representing background knowledge as structured graphs with defined node and edge attributes, and by constraining semantic representation vectors to specific hyperspherical parameter spaces. This parameterization enables automated alignment through mathematical operations rather than complex procedural updates, improving reliability while managing operational complexity through standardized parameter transformations.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent introduces semantic representation vectors as intermediaries between raw background knowledge and communication processes. These vectors serve as a mediating layer that can be independently aligned and updated, simplifying the operation of knowledge integration while ensuring reliable semantic interpretation through the structured intermediary representation.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Measurement precision

If semantic representation vectors are aligned and uniformed on unit hypersphere, then semantic communication accuracy improves, but the computational complexity and processing requirements increase

Engineering Contradiction:
Improvesemantic representation accuracyVSAvoidcomputational processing requirements
Core Design Contradiction:
Measurement precisionVSPower

Solution Approach 1:

The patent applies spheroidality by constraining semantic representation vectors to reside on unit hyperspheres rather than in unconstrained vector spaces. This geometric constraint provides a natural framework for measuring semantic similarity through angular distances and enables efficient alignment operations using spherical geometry, improving measurement precision while leveraging mathematical properties to manage computational complexity.

Inventive Principle:
Principle #14Spheroidality (Curvature)

Data Source

PatentEP4664355A1Apparatus and method for performing background knowledge update on basis of semantic representation in semantic communication
Publication Date: 2025.12.17 LG ELECTRONICS INC
  • EP4664355A1 patent drawingFigure 1
  • EP4664355A1 patent drawingFigure 2
  • EP4664355A1 patent drawingFigure 3

AI summary

According to various embodiments of the present disclosure, there is provided an operating method of a first node in a communication system. The method includes receiving one or more synchronization signals from a second node; receiving system information from the second node; transmitting a random access preamble to the second node; receiving a random access response from the second node; receiving a request for capability information of the first node from the second node; transmitting the capability information to the second node; receiving, from the second node, configuration information for constructing a relation between a background knowledge graph and a semantic representation vector based on the capability information, wherein the configuration information includes information about a second number that is a reference size of the background knowledge graph; and performing construction of the relation through connections of edges between the semantic representation vector and first node vectors in the background knowledge graph based on the configuration information and a first number of nodes by which the first node configures the background knowledge graph. The first node vectors include a selected few of second node vectors configuring the background knowledge graph.