Geographic Tile Matching Nodes for IoT Subscription Scalability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing location-dependent publish and subscribe systems face scalability issues due to reliance on classic indexing data structures, leading to potential service interruptions and inability to handle the exponential growth of IoT data.
Innovation Solution
A horizontally scalable method and system using a grid layout with matching nodes that detect overlaps between subscriptions and publications within defined tiles, employing consistent or rendezvous hashing to distribute computing load and avoid redundant message triggers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a central server is used for processing all matching operations, then the system is simple to manage, but the system lacks fault tolerance and cannot handle exponential growth of IoT data
Solution Approach 1:
The patent divides the geographical region into multiple tiles and assigns each tile to a different matching node. This segmentation distributes the processing workload across multiple independent nodes, eliminating the single point of failure in central server architectures while maintaining manageable system structure through clear division of responsibilities.
Solution Approach 2:
The patent transitions from a centralized vertical architecture to a distributed horizontal architecture by introducing the spatial dimension of geographical tiles. Each matching node operates independently within its assigned tile, creating a two-dimensional distribution pattern that enhances fault tolerance while preserving system manageability through spatial organization.
2Productivity
If multiple matching nodes are added to improve scalability, then the system can handle more data, but message duplication occurs when matches are detected in multiple tiles
Solution Approach 1:
The patent establishes an a priori agreement among matching nodes before matches occur. Each node is pre-configured with rules to determine whether it should trigger notifications for matches involving its tile, based on the tile's position in the geographical hierarchy. This preliminary arrangement prevents message duplication by ensuring only the appropriate node triggers notifications, while maintaining the scalability benefits of multiple matching nodes.
3Ease of manufacture
If classic indexing data structures are used for matching, then the implementation is straightforward, but the system cannot scale horizontally to handle exponential growth
Solution Approach 1:
The patent segments the geographical space into tiles and distributes them across multiple matching nodes, enabling horizontal scalability. Each node maintains a simplified index for its assigned tiles, keeping implementation straightforward while the distributed architecture provides the versatility to scale horizontally to handle exponential growth of IoT data.
Solution Approach 2:
The patent introduces the spatial dimension of geographical tile distribution, transforming the system from a single-dimensional centralized index to a multi-dimensional distributed structure. This enables horizontal scalability across multiple nodes while maintaining implementation simplicity through consistent application of the tile-based segmentation model.
4Productivity
If a vertically scalable architecture is used, then the central server can handle increased load, but the system becomes a single point of failure and cannot maintain service during maintenance
Solution Approach 1:
The patent segments the processing capacity across multiple matching nodes, each handling specific geographical tiles. This distribution maintains high processing capacity through parallel operation while eliminating the single point of failure, as each node can continue serving its tiles independently during maintenance or failure of other nodes.
Solution Approach 2:
The patent transitions from vertical scaling (adding resources to a single server) to horizontal scaling (adding independent nodes across a distributed network). This dimensional shift enables both high processing capacity through parallel processing and service continuity through independent node operation, allowing maintenance without service interruption.
Data Source
AI summary
A method for matching subscriptions of subscribing client nodes (5) with publications of publishing client nodes (4),wherein each subscription is associated with a territory of the subscription (2B) and wherein each publication is associated with a territory of the publication (2A);wherein a geographical region (7) is divided into a plurality of tiles (1);wherein each tile (1) has an associated matching node (3) for detecting matches between publications and subscriptions in that tile;wherein a match is detected when the territory associated with a given subscription and the territory associated with a given publication overlap in that tile (1);wherein the matching node is arranged to trigger an action such as for example notifying the subscribing client node when such a match has been detected;wherein each matching node (3) is further arranged to detect when a match between a particular publication and a particular subscription will be detected in more than one tile (1);and wherein said matching node (3) uses an a priori agreement for determining whether it triggers said action or whether another matching node (3) will do it when a match between a particular publication and a particular subscription will also be detected by the other matching node.