Mobile Neural Intent Translation for Network Slice Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network slice selection methods struggle to accurately translate applications' intents due to mismatched vocabulary sets between applications and networks, limiting the ability to provide tailored network services.
Innovation Solution
Utilizing neural networks on mobile devices and network nodes to translate application intents using different vocabulary sets, enabling efficient network slice selection through reinforcement learning and feedback mechanisms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional network slice selection methods are used, then the network can provide basic services, but it cannot accurately translate application intents due to vocabulary mismatch
Solution Approach 1:
The patent introduces neural networks as intermediary components that translate between application intents and network slice selections. The mobile device neural network translates application intents into network-understandable messages, while the network node neural network translates received messages back into satisfaction evaluations. These neural networks act as mediators that bridge the vocabulary gap between applications and the network, enabling accurate intent translation without requiring direct vocabulary compatibility.
Solution Approach 2:
The patent changes the parameter representation by using learned embeddings and neural network transformations instead of direct vocabulary matching. The system transforms discrete vocabulary items into continuous vector representations that can be processed by neural networks, allowing for nuanced intent translation that captures semantic meaning rather than relying on exact vocabulary matches.
2Measurement precision
If neural networks with reinforcement learning are deployed, then intent translation accuracy improves, but device and network complexity increases
Solution Approach 1:
The patent segments the neural network functionality into two separate components: a mobile device neural network for translating application intents into network messages, and a network node neural network for translating messages back into satisfaction evaluations. This segmentation distributes the computational complexity across different entities, preventing any single device from bearing the full burden of the translation system while maintaining overall accuracy.
Solution Approach 2:
The reinforcement learning mechanism enables the neural networks to automatically improve their translation accuracy through feedback from the network slice satisfaction evaluations. The system self-adjusts by learning from past translations and their outcomes, reducing the need for manual configuration and tuning of the complex neural network parameters while maintaining high slice selection accuracy.
3Ease of operation
If vocabulary sets are standardized across all applications, then network management becomes easier, but applications lose the ability to express specific intent nuances
Solution Approach 1:
The patent transforms the vocabulary representation from discrete standardized terms into continuous vector embeddings that capture semantic nuances. This parameter transformation allows applications to express specific intent nuances through varied embeddings while the neural network learns to map these diverse representations to appropriate network slice selections, maintaining both information precision and management simplicity.
Solution Approach 2:
The neural network acts as an intermediary that receives diverse application intents expressed in various vocabularies and translates them into a standardized internal representation for network slice selection. This intermediary layer preserves the richness of application-specific intent expressions while presenting a unified interface to the network management system, preventing information loss during translation.
Data Source
AI summary
There is provided a method performed by a mobile device, the method comprising: receiving, from an application running on the mobile device, an input defining an intent of the application, the intent being expressed using a first vocabulary set; generating, using a first neural network and based on the intent, a message for requesting a network slice selection for the application, the message being expressed using a second vocabulary set; sending the generated message to a network node; receiving, from the network node, information indicating a network slice selected for use by the application; determining whether the selected network slice satisfies the intent; and updating the first neural network based on the result of the determining.