Communication Network Knowledge Repository for Transfer Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current communication systems lack support for Transfer Learning, making it difficult to develop and deploy new Machine Learning (ML) entities efficiently, as they do not provide means to leverage shared knowledge between similar domains and tasks, hindering the reuse of knowledge from existing ML entities.
Innovation Solution
A centralized knowledge repository and methods for Transfer Learning are introduced, allowing ML entities to register and share knowledge, enabling consumers to request and adapt existing knowledge for new domains, tasks, or network problems, facilitating the reuse of knowledge through meta-description-based searching and sharing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If Transfer Learning is not supported in current communication systems, then developing new ML entities requires huge amounts of data and computational consumption, but with Transfer Learning support, knowledge can be reused from existing ML entities
Solution Approach 1:
The patent implements a centralized knowledge repository that stores pre-acquired knowledge from existing ML entities. This preliminary action allows new ML entities to leverage previously acquired knowledge, eliminating the need to start from scratch and significantly reducing development time and computational requirements.
Solution Approach 2:
The patent enables copying of knowledge from source ML entities to target ML entities through the centralized repository. This copying mechanism allows the reuse of trained models, features, and parameters, dramatically reducing the time and resources needed for developing new ML entities while maintaining high productivity.
2Use of energy by moving object
If knowledge sharing between ML entities is not enabled, then each ML entity must be developed independently with full computational resources, but with knowledge sharing, the system complexity increases
Solution Approach 1:
The patent introduces a centralized knowledge repository as an intermediary between source and target ML entities. This mediator manages knowledge storage, retrieval, and sharing, reducing computational consumption for individual ML entities while centralizing the complexity of knowledge management, thus balancing energy efficiency with system complexity.
3Adaptability or versatility
If ML entities cannot adapt existing knowledge to new domains, then knowledge reuse is limited to identical contexts, but adapting knowledge requires additional processing
Solution Approach 1:
The patent implements dynamic knowledge adaptation where the centralized repository and ML entities can adjust knowledge based on domain similarity assessments. This dynamic approach allows knowledge to be flexibly adapted to new domains through automated similarity evaluation and selective adaptation, enhancing versatility while managing adaptation complexity through intelligent algorithms.
Data Source
AI summary
A management service producer or data repository of a communication network provides at least one of the following: information about knowledge available for sharing, or knowledge available for sharing, and executes a transfer learning process for sharing at least part of the knowledge. A management service consumer of the communication network generates at least one of the following: a request for information on available knowledge from the management service producer or data repository, a request for available knowledge from the management service producer or data repository, or a request for executing a transfer learning process between the management service producer or data repository and an ML entity or ML-enabled function of the communication network, andmanages the request for available knowledge or the request for executing the transfer learning process, and/or adapts content of the request for available knowledge or the request for executing the transfer learning process.