HSDPA Scheduler Dividing Priority Queues
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing HSDPA scheduling technologies face limitations in managing end-user traffic efficiently due to hardware resource constraints, as all priority queues of users are considered for transmission in each sub-frame, leading to high clock cycle capacity and memory consumption.
Innovation Solution
The proposed solution involves a scheduler unit that allocates resources efficiently by dividing priority queues into two subsets, where only the second subset is considered for scheduling at each transmission opportunity, utilizing fast and slow access memory partitions to optimize resource allocation and reduce CPU load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If all priority queues of all users are considered for transmission in each sub-frame, then scheduling comprehensiveness is improved, but clock cycle capacity and memory consumption increase
Solution Approach 1:
The patent segments the set of all priority queues into two distinct subsets: an active subset and an inactive subset. The scheduler only considers queues in the active subset for transmission decisions, while queues in the inactive subset are temporarily excluded. This segmentation reduces the number of queues evaluated during each scheduling decision, thereby reducing clock cycle consumption and memory access requirements while maintaining scheduling effectiveness through periodic reactivation of inactive queues.
2Reliability
If all priority queues are evaluated in each scheduling instant, then scheduling fairness is improved, but CPU load increases
Solution Approach 1:
The patent implements a dynamic subset management mechanism where the composition of the active priority queue subset changes over time based on scheduling needs and system conditions. Queues are dynamically transferred between the active and inactive subsets, allowing the scheduler to adapt its evaluation scope. This dynamic adjustment maintains scheduling fairness by ensuring all queues eventually receive consideration while reducing instantaneous CPU load by limiting the number of queues evaluated at any given moment.
Data Source
AI summary
A first interface for receiving downlink data packets from a core network, a second interface for transmitting data packets to user entities, a processor comprising a scheduler, a first memory partition comprising and parameter data pertaining to a first subset of allocated priority queues, a second memory partition comprising parameter data pertaining to a second subset of allocated priority queues, is provided. The processor is transferring parameter data between the first and second memory partitions, and vice versa, whereby parameter data for priority queues having no data packets in corresponding buffers is prioritized for being transferred to the first memory partition, parameter data for priority queues having data packets in corresponding buffers is prioritized for being transferred to the second memory partition, the processor performing scheduling based on a selection of priority queues whose parameter data is residing in the second memory partition.