Scheduler-Managed Cloud-RAN Packet Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The distribution of Radio Access Network (RAN) processing within a data center in a centralized radio access network (CRAN) poses challenges in minimizing delay and optimizing compute resource utilization, as each user requires independent processing, and existing methods fail to efficiently manage the distribution of RAN processing tasks across multiple compute resources.
Innovation Solution
The method involves assigning RAN identification information to packets from remote radio access points, which is used by schedulers to determine the compute resources for processing and routing within the data center, utilizing a hierarchical scheduling system integrated with the network switching fabric, allowing for implicit load balancing and reduced overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If RAN processing is centralized in a data center with multiple compute resources, then processing capacity and scalability are improved, but routing complexity and delay increase due to the need to distribute packets across multiple compute resources
Solution Approach 1:
The patent introduces RAN identification information as an intermediary element that mediates between packets and compute resources. This identifier acts as a simplified interface that replaces complex routing tables and addressing schemes, enabling direct packet routing to compute resources based on the RAN ID without requiring complex network infrastructure or multiple routing decisions.
Solution Approach 2:
The patent extracts the routing decision logic from the network switching fabric and consolidates it into schedulers that use RAN identification information. By separating the identification function from the routing function, the system simplifies the routing process while maintaining the ability to distribute packets across multiple compute resources efficiently.
2Measurement precision
If packets are routed using explicit compute resource addressing, then routing precision is improved, but overhead and processing delay increase due to complex addressing and scheduling
Solution Approach 1:
The patent changes the addressing parameter from complex compute resource identifiers to simple RAN identification information. This parameter transformation maintains routing precision by uniquely identifying the target compute resource through the RAN ID, while significantly reducing the time required for routing decisions and packet processing.
3Reliability
If independent processing is provided for each user, then processing reliability is improved, but resource utilization efficiency deteriorates due to lack of load balancing
Solution Approach 1:
The patent implements dynamic load balancing through schedulers that continuously monitor compute resource status and dynamically assign packets to appropriate compute resources based on current load conditions. This dynamic approach maintains independent user processing for reliability while optimizing resource utilization by distributing workloads adaptively across available compute resources.
Solution Approach 2:
The system incorporates feedback mechanisms where schedulers receive information about compute resource status and use this feedback to make intelligent routing decisions. The feedback loop enables the system to balance individual user processing requirements with overall resource utilization efficiency by adjusting packet distribution based on real-time conditions.
Data Source
AI summary
A method of managing Radio Access Network (RAN) processing within a data center of a centralized radio access network (CRAN), wherein the data center comprises a plurality of compute resources that perform at least part of the RAN processing of a plurality of connected remote radio access points, includes assigning, to packets from the plurality of remote radio access points, an identification information that originates from the RAN; deciding, by schedulers based on a scheduling algorithm, on which of the compute resources processing of the packets is to take place; and using the identification information to route the packets within the data center. The compute resources are implicitly addressed through the identification information.