Coordinate-Based Indexing for Edge Computing Data Retrieval
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data indexing mechanisms in edge computing face challenges such as high bandwidth costs, performance bottlenecks, and increased latency due to inefficient data retrieval processes, particularly in distributed environments where data is cached across geographically diverse edge servers.
Innovation Solution
A coordinate-based indexing (COIN) mechanism that utilizes Software Defined Networking (SDN) to reduce data indexing costs and latency by determining switch coordinates in a virtual space, allowing for direct forwarding of data index requests to the nearest index server, thereby minimizing the number of forwarding entries and path lengths.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a full index mechanism is used where each edge node maintains a complete index of all data in the edge network, then data retrieval speed is improved, but bandwidth cost and system complexity increase significantly
Solution Approach 1:
The patent segments the indexing system into two distinct components: index publishing nodes that maintain data location information and index query nodes that search for data. This segmentation allows the system to achieve fast data retrieval without requiring every node to maintain a complete index, thereby reducing overall system complexity while preserving retrieval speed.
Solution Approach 2:
The patent introduces an index server as an intermediary component that centralizes index management. Instead of each edge node maintaining complete indexes, the index server acts as a mediator that stores and manages all index information, allowing query nodes to efficiently retrieve data locations without the complexity of distributed full indexing.
2Device complexity
If a centralized index server is used to store all data indexes, then indexing complexity is reduced, but performance bottleneck and fault tolerance issues arise
Solution Approach 1:
The patent segments the centralized index server into multiple distributed index servers. Each index server maintains a portion of the index space, distributing the storage and query processing load across multiple nodes. This segmentation eliminates the single point of bottleneck while maintaining manageable complexity through modular architecture.
Solution Approach 2:
The patent introduces a hierarchical dimension to the indexing system, organizing index servers in a distributed hierarchy where index information is partitioned across multiple levels. This dimensional organization allows the system to scale horizontally, improving throughput by adding more index servers without proportionally increasing complexity at any single level.
3Device complexity
If DHT indexing mechanism is used in peer-to-peer networks, then distributed storage is achieved, but query processing latency increases due to multiple overlay hops
Solution Approach 1:
The patent applies preliminary action by pre-computing and caching index locations in a distributed manner before queries are issued. Index publishing nodes proactively publish data location information to the distributed index structure in advance, so when queries arrive, the system can quickly retrieve pre-computed index information without performing extensive multi-hop searches, thereby reducing query latency while maintaining distributed indexing capabilities.
4Reliability
If data is retrieved from cloud for every edge user request, then data availability is ensured, but network bandwidth consumption and delay increase
Solution Approach 1:
The patent applies preliminary action through edge caching, where frequently accessed data is pre-fetched and cached at edge servers before actual user requests. This allows the system to serve edge users from local caches rather than retrieving data from the cloud for every request, significantly reducing network bandwidth consumption while maintaining data availability through the cache hierarchy.
Solution Approach 2:
The patent implements local quality by enabling each edge server to maintain local caches of frequently accessed data. This allows data to be served from the nearest edge server with appropriate cache hits, reducing the need for remote cloud retrieval and thereby decreasing network bandwidth consumption while ensuring data availability through localized storage.
Data Source
AI summary
The invention discloses an implementation method of unstructured data sharing mechanism for edge computing and a system thereof, wherein the method includes the following steps: receive a data index request from a user, collect the switch, port, link and host information through the control plane to obtain the network topology and status; determine the coordinates of the switch and the coordinates of the data index in the virtual space formed by the network topology; construct a switch connection diagram according to the coordinates of the switch; insert the coordinates of the switch and forwarding entries of the switch connection diagram in the forwarding table of the switch; the data plane greedily forwards the coordinates of the data index until being transmitted to the switch corresponding to the switch coordinates closest to the coordinates of the data index in the virtual space.