HFL Subscription Control for Dynamic AIMLE Training Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Horizontal Federated Learning (HFL) capabilities are limited to subscribing to AIMLE clients associated with an HFL training event.
Innovation Solution
The introduction of additional functionalities for an HFL training service, including the ability to modify or retrieve information related to an HFL training service and to terminate an HFL subscription, enhancing the training process by enabling the AIMLE server to have greater control over the training process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the AIMLE server subscribes to AIMLE clients for HFL training events, then the server can obtain ML model parameters, but the server lacks control to modify or terminate subscriptions
Solution Approach 1:
The subscription mechanism is transformed from static (once subscribed, cannot be modified or terminated) to dynamic (can be modified or terminated at any time). The server can now dynamically adjust subscription parameters such as training event identifiers, client identifiers, and subscription identifiers, allowing flexible control over the HFL training process while maintaining manageable complexity through standardized procedures.
2Productivity
If the HFL training service uses basic subscription functionality, then the system remains simple, but the training process lacks flexibility and efficiency
Solution Approach 1:
The server performs preliminary actions by establishing detailed subscription information before HFL training events occur. Subscription identifiers, client identifiers, and training event identifiers are pre-configured and stored, allowing the server to efficiently manage multiple training operations without ad-hoc configuration, thereby improving productivity while maintaining ease of operation through automated reference to stored subscription data.
Data Source
AI summary
Various aspects of the present disclosure relate to a User Equipment (UE) configured to or operable to receive, from a network entity, a request associated with a Horizontal Federated Learning (HFL) subscription for retrieving information associated with the HFL subscription, for modifying information associated with the HFL subscription, or for terminating the HFL subscription, and transmit, to the network entity, a response associated with the HFL subscription based at least in part on the received request, wherein the UE is an Artificial Intelligence Machine Learning Enablement (AIMLE) client, and wherein the network entity is an AIMLE server.