Dynamic Resource Allocation in Cloud Radio Access Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud radio access networks face challenges in optimizing resource management due to dynamic user needs, leading to inefficiencies in resource allocation and increased costs, as existing methods struggle to adapt to varying traffic patterns and user mobility.
Innovation Solution
A dynamic resource allocation method using deep reinforcement learning to predict load fluctuations and adjust computational resources in a cloud radio access network, allowing for efficient allocation of resources based on real-time user demands by learning load patterns and optimizing resource distribution across virtual machines.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If resource management is operated for a long time in C-RAN, then system stability is improved, but adaptability to dynamic user needs deteriorates
Solution Approach 1:
The patent implements dynamic resource allocation by separating the baseband unit into virtual machines that can be dynamically created, migrated, and scaled based on real-time traffic conditions. The resource allocation is no longer static but continuously adjusted according to user demand patterns, enabling the system to adapt while maintaining operational stability through controlled virtualization.
Solution Approach 2:
The system changes key operational parameters such as computational resource allocation, virtual machine instantiation, and baseband processing capacity based on monitored traffic conditions. By dynamically adjusting these parameters rather than maintaining fixed configurations, the system achieves both stability through controlled changes and adaptability to varying user needs.
2Adaptability or versatility
If computational resources are increased to meet diverse user needs, then adaptability is improved, but resource efficiency deteriorates
Solution Approach 1:
Instead of allocating full computational resources to all virtual machines continuously, the system applies partial resource allocation based on actual traffic demand. Virtual machines receive computational resources proportionally to their current load requirements, avoiding excessive resource consumption while maintaining the capability to scale up when needed, thus improving both adaptability and efficiency.
Solution Approach 2:
The virtualized baseband unit pool creates universal computational resources that can serve multiple functions and multiple virtual machines simultaneously. A single physical resource pool supports diverse user needs through virtualization, eliminating the need for dedicated resources for each function and improving overall resource efficiency while maintaining adaptability.
3Loss of energy
If deep reinforcement learning is used for resource allocation, then resource efficiency is improved, but device complexity increases
Solution Approach 1:
The patent introduces a deep reinforcement learning-based resource allocation unit as an intermediary layer between the virtual machines and the physical computational resources. This intermediary intelligently manages resource distribution by learning from traffic patterns and making optimized allocation decisions, improving resource efficiency while isolating the complexity of the learning algorithm from the core baseband processing functions.
Data Source
AI summary
Disclosed is an apparatus for dynamic resource allocation in cloud radio access networks, and the dynamic resource allocation apparatus includes: a deep reinforcement learning unit learning load fluctuation of a remote radio head by using deep reinforcement learning and predicting the load fluctuation of the remote radio head; a calculation unit calculating a computational resource of a virtual machine corresponding to the remote radio head by using the predicted load fluctuation; and an allocation unit allocating the calculated computational resource to the virtual machine.