Distributed Cache Management for Low-Latency Information-Centric Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current TCP/IP-based content distribution architectures and ICN models with on-path caching suffer from inefficiencies such as high latency, low cache hit ratio, and increased overhead due to reliance on centralized control, leading to suboptimal data transmission in information-centric networks.
Innovation Solution
A distributed caching and request routing method that combines on-path and off-path caching techniques, utilizing broadcast look-ups among neighbor nodes within a limited hop range to optimize cache utilization and reduce latency, based on node properties and data popularity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If on-path caching is used, then cache management is simple, but cache hit ratio is low and latency is high
Solution Approach 1:
The patent segments the caching system into two independent parts: on-path caches (for simple management) and off-path caches (for improved hit ratio). Each part operates autonomously with its own cache manager, allowing the system to benefit from both simplicity and enhanced performance without requiring complete redesign of either component.
Solution Approach 2:
The patent introduces a content requestor as an intermediary component that coordinates between on-path and off-path caches. This mediator handles the complexity of searching multiple cache locations (on-path nodes and off-path neighbors) and managing the caching decisions, thereby enabling improved cache hit ratio without increasing overall system complexity.
2Reliability
If centralized control is used, then coordination is improved, but overhead and bottlenecks increase
Solution Approach 1:
The patent implements self-service by enabling cache managers at different network elements (on-path nodes and off-path nodes) to autonomously manage their own caches without centralized control. Each cache manager independently handles cache insertion, removal, and search operations, eliminating the overhead and bottlenecks associated with centralized coordination while maintaining effective coordination through distributed protocols.
3Reliability
If off-path look-up is performed, then cache hit ratio is improved, but search scope and overhead increase
Solution Approach 1:
The patent applies local quality by limiting off-path look-up to a localized scope around the current on-path node (specifically to off-path neighbors within one hop). This localized approach improves cache hit ratio by checking nearby nodes while avoiding the excessive overhead and complexity of searching the entire network. The solution strikes a balance between search thoroughness and computational cost.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a computer-implemented method for transmitting data (18) in an information-centric network (12), the information-centric network (12) being connectable to a plurality of server devices (14, 14a-f) for providing network service data (18), respectively, the information-centric network (12) being configured for receiving a request (24) for network service data (18) from a client device (16, 16a-f), wherein the information-centric network (12) includes a plurality of interconnected nodes (20) including, respectively, a cache for caching network service data (18), each of the nodes (20) being configured for routing any request (24) to another of the nodes (20) along a corresponding routing path (26) from the client device (16, 16a-f) to the corresponding server device (14, 14a-f), the corresponding routing path (26) having one or more nodes (28, 28a-g) successively visitable by the request (24) via one hop.