Buffering Scheduling Parameter Changes in Resource-Constrained Node Bs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Resource-limited HSPA Node Bs face challenges in processing a high number of scheduling priority indicator (SPI) changes, leading to latency and quality degradation due to hardware limitations, making it difficult to enforce SPI changes and maintain user experience.
Innovation Solution
A mechanism that stores parameter change information independently of immediate processing, allowing for reliable handling of multiple SPI changes by separating the detection and update processes, ensuring that SPI changes are executed based on available processing capacity and maintaining up-to-date information for the MAC-hs packet scheduler.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the Node B processes a high number of SPI changes immediately, then the QoS differentiation is maintained, but the hardware resources are overwhelmed causing latency and processing failures
Solution Approach 1:
The patent stores SPI change information in advance in a buffer memory before actual processing occurs. This preliminary storage allows the system to capture all parameter changes without immediate processing, preventing loss of QoS updates while avoiding overwhelming the processor. The buffered information is processed later when processing capacity is available.
Solution Approach 2:
The patent introduces an intermediate buffer memory structure that decouples the reception of SPI change information from its processing. This intermediary component acts as a buffer between the input stream of parameter changes and the processing unit, allowing the system to handle high volumes of changes without direct processing pressure on the hardware resources.
2Speed
If the Node B updates SPI parameters immediately upon receipt, then real-time QoS adjustment is achieved, but processing latency increases under high traffic conditions
Solution Approach 1:
The system performs preliminary storage of SPI change information in buffer memory without immediate processing. This allows the system to capture all changes at their arrival time, then process them in batches when processing capacity is available, reducing the time each individual parameter update waits for processing while maintaining accurate QoS state.
Solution Approach 2:
The patent implements periodic processing of buffered SPI change information based on processing capacity availability. Instead of attempting continuous immediate processing, the system periodically updates parameters when resources are available, reducing latency accumulation while ensuring all changes are eventually applied.
3Reliability
If the Node B processes all SPI changes with high priority, then QoS enforcement is strict, but other critical processing tasks suffer from resource starvation
Solution Approach 1:
The patent stores SPI change information in advance without immediate processing, creating a buffer that allows the system to maintain QoS parameter state accurately while deferring actual processing. This preliminary storage ensures no QoS updates are lost, while the deferred processing allows other critical tasks to share processor resources fairly.
Solution Approach 2:
The system dynamically adjusts processing timing based on available processing capacity. When processor resources are available, buffered SPI changes are processed; when resources are constrained by other tasks, processing is deferred. This dynamic approach ensures QoS enforcement without starving other critical processing functions.
Data Source
AI summary
An apparatus and a method are described, by which at least one bearer containing data for a user equipment is received and a connection to the user equipment is provided via a radio access network. Furthermore, transmission of the data for the user equipment is scheduled based on a scheduling parameter associated to the bearer. Moreover, parameter change information associated to the bearer is detected, the parameter change information indicating how the scheduling parameter is to be changed, and the parameter change information for the bearer is stored in the memory.