AI Engine Downlink Radio Resource Scheduling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional LTE systems experience long delays and resource wastage due to redundant information processing during downlink resource scheduling, requiring multiple confirmations and feedbacks.
Innovation Solution
An AI engine-supported downlink radio resource scheduling method utilizing a deep reinforcement learning algorithm, specifically the DDQN algorithm, to replace traditional Round-Robin and fair Round-Robin scheduling algorithms, optimizing resource allocation by learning from ever-changing user service demands and minimizing differences in action value distributions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If conventional LTE scheduling algorithms (Round-Robin, fair Round-Robin) are used, then the system structure is simple and easy to implement, but long delays and resource wastage occur due to redundant information processing and multiple feedback loops
Solution Approach 1:
The patent replaces the traditional mechanical/symbolic scheduling algorithms (Round-Robin, fair Round-Robin) with an AI-based deep reinforcement learning model. The AI model processes scheduling decisions through neural network computations, substituting the conventional iterative feedback mechanism with a data-driven approach that directly outputs optimal resource allocation, thereby eliminating redundant processing steps and reducing scheduling delays.
Solution Approach 2:
The patent changes the fundamental parameters of the scheduling system by introducing deep reinforcement learning components (policy network, value network, Q-values) instead of traditional scheduling parameters (time slots, priority queues). This parameter transformation enables the system to make intelligent scheduling decisions based on learned patterns from historical data, achieving faster and more efficient resource allocation without the delays inherent in conventional feedback-based mechanisms.
2Ease of manufacture
If conventional LTE scheduling algorithms are used, then the implementation is straightforward, but a lot of processing delays and resource wastage occur during interactive transmission
Solution Approach 1:
The patent substitutes the straightforward but inefficient conventional scheduling implementation with an AI-enhanced system. The deep reinforcement learning model processes scheduling requests through neural network inference, replacing the mechanical iterative confirmation process with a single-step intelligent decision-making approach, thereby significantly improving resource allocation efficiency while maintaining implementation feasibility through modular AI integration.
Solution Approach 2:
The AI scheduling system performs self-service by automatically learning from historical scheduling data and making optimal decisions autonomously. The deep reinforcement learning model continuously refines its policy through training on past performance, enabling it to handle new scheduling requests efficiently without requiring manual intervention or extensive feedback loops, thus improving productivity while keeping the system relatively simple to implement.
3Reliability
If multiple confirmations and feedback loops are used in LTE, then information transmission reliability is improved, but long delays are produced and resources are wasted
Solution Approach 1:
The patent incorporates feedback mechanisms within the AI model itself, where the deep reinforcement learning algorithm uses historical scheduling data and outcome information to continuously improve its decision-making policy. This internal feedback loop replaces the external iterative confirmation process, allowing the system to learn from past performance and make more reliable scheduling decisions faster, thereby maintaining reliability while reducing feedback-related delays.
Solution Approach 2:
The AI model performs preliminary learning and decision-making based on predicted outcomes before actual resource allocation occurs. By training on historical data and simulating scheduling scenarios in advance, the system prepares optimal allocation strategies beforehand, reducing the need for multiple real-time confirmations and feedback loops, thus maintaining information transmission reliability while significantly cutting down on feedback loop delays.
Data Source
AI summary
An Artificial Intelligence (AI) engine-supporting downlink radio resource scheduling method and apparatus are provided. The AI engine-supporting downlink radio resource scheduling method includes: constructing an AI engine, establishing a Socket connection between an AI engine and an Open Air Interface (OAI) system, and configuring the AI engine into an OAI running environment to utilize the AI engine to replace a Round-Robin scheduling algorithm and a fair Round-Robin scheduling algorithm adopted by a Long Term Evolution (LTE) at a Media Access Control (MAC) layer in the OAI system for resource scheduling to take over a downlink radio resource scheduling process; sending scheduling information to the AI engine through Socket during the downlink radio resource scheduling process of the OAI system; and utilizing the AI engine to carry out resource allocation according to the scheduling information, and returning a resource allocation result to the OAI system.