Communication Network Parameter Orchestration via ML Agent Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for optimizing operational parameters in communication networks, such as RET and maximum DL transmit power, often fail to account for the interdependencies between these parameters and their impact on network Key Performance Indicators (KPIs), leading to suboptimal decisions.
Innovation Solution
A computer-implemented method and orchestration node that use Machine Learning (ML) to predict which operational parameter agent should execute an action to maximize a performance measure for the communication network, thereby coordinating the management of multiple operational parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple independent agents optimize different operational parameters (RET, maximum DL transmit power, PO Nominal PUSCH) separately, then each parameter can be optimized individually, but the interdependencies between parameters and their combined impact on network KPIs are not accounted for, leading to suboptimal overall network performance
Solution Approach 1:
The patent merges multiple independent parameter optimization agents into a unified hierarchical orchestration system. The higher-level orchestrator coordinates the lower-level parameter agents (RET, maximum DL transmit power, PO Nominal PUSCH) to consider their interdependencies, ensuring that optimization decisions account for combined impacts on network KPIs rather than treating each parameter in isolation.
Solution Approach 2:
The patent introduces a higher-level orchestrator as an intermediary between the lower-level parameter agents and the network KPIs. This orchestrator mediates the optimization process by evaluating the combined effects of multiple parameter changes and coordinating agent actions to achieve optimal overall network performance, preventing suboptimal decisions that would arise from independent optimization.
2Reliability
If coordinate optimization of multiple parameters is implemented to account for interdependencies, then overall network performance can be improved, but the system complexity and coordination overhead increase significantly
Solution Approach 1:
The patent segments the optimization system into distinct hierarchical levels: a higher-level orchestrator and lower-level parameter-specific agents. This segmentation allows the complex coordination task to be divided into manageable components, where each agent focuses on its specific parameter while the orchestrator handles the coordination, reducing overall system complexity compared to a fully centralized approach.
Solution Approach 2:
The patent implements dynamic coordination where the higher-level orchestrator adaptively manages the lower-level agents based on current network conditions. This dynamic approach allows the system to adjust its coordination strategy in real-time, optimizing performance while managing complexity through flexible, condition-based decision-making rather than rigid fixed-structure coordination.
3Productivity
If independent parameter optimization is performed without considering other parameters, then the optimization process is simpler and faster, but the decisions made by one agent may conflict with or undermine the optimization efforts of other agents
Solution Approach 1:
The patent implements preliminary coordination at the higher-level orchestrator before lower-level agents execute their optimization actions. The orchestrator evaluates the current state and planned actions of multiple agents, ensuring that their combined effects will be beneficial before allowing execution. This preliminary coordination prevents conflicting actions while maintaining relatively fast optimization throughput.
Solution Approach 2:
The patent implements feedback mechanisms where the higher-level orchestrator monitors the effects of parameter changes on network KPIs and uses this information to coordinate subsequent agent actions. This feedback loop ensures that optimization decisions are effective and consistent, preventing one agent's actions from undermining another's optimization efforts while maintaining efficient optimization pace.
Data Source
AI summary
A method (200) is disclosed for orchestrating management of a plurality of operational parameters in an environment of a communication network. Each of the operational parameters is managed by a respective Agent, and at least one performance parameter of the communication network is operable to be impacted by each of the operational parameters. The method comprises obtaining a representation of a state of the environment (210), and generating a prediction, using an ML process and the obtained state representation, of which of the Agents, if allowed to execute within the environment an action selected by the Agent for management of its operational parameter, will result in the greatest increase of a performance measure for the communication network (220). The method further comprises selecting an Agent on the basis of the prediction (230) and initiating execution by the selected Agent of its selected action (240).