Hierarchical Cluster Tree Overlay Network Topology Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing topology-aware overlay network protocols fail to accurately reflect underlying physical network topology, leading to inefficient data placement and routing due to mismatches between logical and physical network structures.
Innovation Solution
The Hierarchical Cluster Tree (HCT) protocol uses network distance measurements to construct a hierarchical clustering structure that groups nodes based on physical proximity, utilizing local area networks (LANs) as building blocks to create a distributed overlay network that closely matches the physical network topology, enabling efficient routing and data placement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If landmark-based protocols use RTT to group nodes, then nodes can be grouped together based on distance from a landmark, but nodes spaced in different directions from the landmark may be grouped together even though they are quite distant from each other
Solution Approach 1:
The patent introduces a synthesized network map as an intermediary data structure that captures the underlying physical topology. Instead of directly grouping nodes based on RTT from a single landmark, the system first builds a comprehensive network map that represents the physical network structure, then uses this map to guide node grouping. This intermediary representation resolves the contradiction by providing more accurate physical proximity information than simple RTT measurements from one landmark.
Solution Approach 2:
The patent changes the parameters used for node grouping from simple RTT values to a more comprehensive set of parameters including RTT measurements to multiple landmarks, bandwidth measurements, and synthesized network map data. By changing these parameters, the system achieves more accurate physical proximity assessment and avoids grouping nodes that are actually distant from each other.
2Loss of information
If synthetic network maps are used to capture underlying physical topology, then physical network structure can be represented, but the techniques suffer from inaccuracies
Solution Approach 1:
The patent performs preliminary actions to improve the accuracy of the synthesized network map before it is used for routing and data placement. This includes collecting RTT measurements to multiple landmarks, performing bandwidth measurements, and using these measurements to refine the network map. By performing these preliminary actions, the system reduces inaccuracies in the topology representation.
Solution Approach 2:
The system uses feedback from multiple RTT measurements and bandwidth measurements to continuously refine and update the synthesized network map. By collecting feedback from multiple sources and using it to adjust the network map representation, the system improves the accuracy of the physical topology capture over time.
3Device complexity
If overlay networks are formed without consideration of underlying physical topology, then routing is based on logical topology only, but routing decisions that are sensible at the logical level may be far from optimal at the physical level
Solution Approach 1:
The patent applies local quality by making routing decisions that are optimized for the local physical network conditions. Instead of using a single global routing strategy, the system uses the synthesized network map to identify local physical topology characteristics and applies routing strategies that are appropriate for each local region. This allows routing to be simple within local regions while being optimal for the physical network overall.
Solution Approach 2:
The patent adds another dimension to routing by incorporating physical topology information from the synthesized network map into the logical routing decisions. The system operates in both the logical overlay dimension and the physical underlay dimension, using information from both dimensions to make routing decisions. This dimensional addition allows the system to maintain logical routing simplicity while achieving physical routing optimality.
Data Source
AI summary
A Hierarchical Cluster Tree (HCT) overlay network reflects underlying physical network topology including inter-node distances (e.g., hop count), and an HCT structure groups nodes based on distance measurements. Local area networks (LANs) are basic building blocks. The network includes first and second level-1 clusters of LANs, each LAN including computerized data processing and/or storage nodes. A distance between LANs in the level-1 clusters is less than or equal to a predefined level-1 maximum distance. The LANs of the first and second level-1 clusters form a level-2 cluster of LANs in which a distance between cluster medoid nodes of the level-1 clusters is less than or equal to a predefined level-2 maximum distance greater than the level-1 maximum distance. Clustering can be extended to higher levels having successively greater maximum distances.