Source-Aggregation Tree Resource Data Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network technologies lack efficient mechanisms for providing quality of service (QoS) control on a per-flow basis, leading to traffic congestion and inefficiencies in access networks due to high algorithm costs and scalability issues with existing methods like ATM and IntServ/RSVP.
Innovation Solution
The method employs Source-Aggregation Trees (SAT) and Interval-Based Upstream Traffic Accounting (UTA) to reduce memory and time costs by aggregating traffic accounting and decoupling traffic registration from accounting processes, using the SAT-RM algorithm to manage resource data efficiently within the network domain.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If per-flow QoS control is implemented using traditional methods like ATM or IntServ/RSVP, then quality of service control capability is improved, but algorithm cost and computational complexity increase significantly
Solution Approach 1:
The patent segments the network into hierarchical domains with domain controllers and member nodes, dividing the complex per-flow QoS control problem into manageable parts. Each domain controller manages QoS for its local domain while maintaining aggregated views, reducing overall computational complexity while preserving QoS control capability.
Solution Approach 2:
The patent introduces domain controllers as intermediary entities between individual flows and the core network. These controllers aggregate traffic accounting information from multiple flows and present summarized data upstream, reducing the algorithmic burden on core network elements while maintaining per-flow QoS control capability within domains.
2Measurement precision
If traffic accounting for all data flows is maintained to enable per-flow QoS control, then measurement precision is improved, but memory cost and scalability worsen
Solution Approach 1:
The patent segments traffic accounting into two levels: detailed per-flow accounting maintained by domain controllers for local QoS decisions, and aggregated accounting maintained by upstream elements for resource management. This segmentation allows precise measurement where needed while reducing overall memory requirements through aggregation.
Solution Approach 2:
The patent applies partial action by maintaining full traffic accounting only for flows within specific domains that require per-flow QoS control, while using aggregated accounting for other traffic. This selective approach reduces memory costs while maintaining measurement precision for critical flows.
3Loss of information
If all nodes are informed of all resource data changes in the network, then information completeness is improved, but communication overhead and processing time increase
Solution Approach 1:
The patent segments the notification process by domain and tree structure. When resource data changes occur, only affected domains and their member nodes receive notifications through targeted updates in their respective SAT trees. This segmentation maintains information completeness for relevant nodes while dramatically reducing communication overhead and notification time.
Solution Approach 2:
The patent applies local quality by providing detailed resource data information to nodes that need it for QoS decisions, while providing aggregated summaries to other nodes. Each node receives information appropriate to its function and location in the hierarchy, reducing unnecessary communication while maintaining local decision-making capability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method of distributing and aggregating resource data, such as transmission capacity. First, for each edge node in the domain a respective spanning tree is constructed that connects said edge node to ail other edge nodes and that identifies for each node other than said edge node in said tree a single upstream neighbour node and zero or more downstream neighbour nodes. Second, any node that has a change in resource data in connection with a particular spanning tree it is a member of, such as the transmission capacity increase or decreases, will send a message to update each of its upstream and downstream neighbour in that spanning tree on said change. Third, each such neighbour will calculate impact of said change to its own resource data and will send a message to update each of its upstream and downstream neighbour in that spanning tree on said impact.