IAB Resource Allocation Using Token Buckets for UE Bearer Fairness
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing IAB networks face issues with fairness between UE bearers due to the lack of granular QoS enforcement in current BAP and MAC layer designs, leading to unfair data transmission and potential QoS degradation.
Innovation Solution
Implement a resource allocation method that enhances fairness by acquiring UE bearer information and allocating resources based on the number of tokens maintained for each UE bearer, using methods such as dividing MAC SDUs into different queues and applying round-robin allocation, ensuring fairness across multiple hops.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If many-to-one mapping is used between UE bearers and BH RLC channels, then resource utilization is improved, but fairness between UE bearers deteriorates
Solution Approach 1:
The patent segments the resource allocation process into multiple stages: first allocating resources to logical channels based on their token buckets, then allocating resources within each logical channel to individual UE bearers based on their token buckets. This segmentation allows the system to maintain fairness at the UE bearer level while enabling efficient resource utilization at the logical channel level, resolving the contradiction between resource utilization and fairness.
Solution Approach 2:
The patent applies different allocation strategies at different levels: logical channels are allocated based on their aggregate token bucket values (prioritizing channels with more available resources), while individual UE bearers within each channel are allocated based on their individual token bucket values. This local quality differentiation ensures that both channel-level efficiency and bearer-level fairness are achieved simultaneously.
2Reliability
If one-to-one mapping is used between UE bearers and BH RLC channels, then fairness between UE bearers is maintained, but resource utilization efficiency deteriorates
Solution Approach 1:
The patent merges multiple UE bearers into logical channels for resource allocation, allowing multiple bearers to share a common channel resource pool. Within each merged channel, individual bearer fairness is maintained through token bucket-based allocation. This merging approach enables better resource utilization by allowing statistical multiplexing while preserving fairness through controlled allocation mechanisms.
Solution Approach 2:
The patent changes the allocation parameter from simple one-to-one mapping to token bucket-based weighted allocation. By introducing token bucket values as the allocation parameter, the system can dynamically adjust resource distribution based on current conditions, achieving both efficiency and fairness. The token bucket mechanism allows the system to transition from rigid one-to-one mapping to flexible many-to-one mapping with fairness guarantees.
3Device complexity
If existing BAP and MAC layer designs are used, then device complexity is reduced, but QoS enforcement granularity deteriorates
Solution Approach 1:
The patent implements a nested structure where UE bearer-level token buckets are contained within logical channel-level token buckets. This nested architecture allows the existing BAP and MAC layers to be enhanced with fine-grained QoS control without completely redesigning the entire system. The nested token bucket mechanism provides granular QoS enforcement while maintaining compatibility with existing protocols, thus balancing complexity and precision.
Solution Approach 2:
The patent introduces token bucket structures as intermediary mechanisms between the existing BAP/MAC layer protocols and the required fine-grained QoS control. These token buckets act as mediators that translate high-level QoS requirements into actionable allocation decisions at the MAC layer, enabling precise QoS enforcement without fundamentally changing the underlying protocol architecture.
Data Source
AI summary
A resource allocation apparatus includes processor circuitry configured to acquire UE bearer information, and processor circuitry configured to allocate a resource for an MAC SDU of a UE bearer k in a logical channel j, allocation of the resource being related to the number of tokens of UE bearer in the logical channel j.