Network-of-Networks Capability Matching for Resource Sharing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing networks do not efficiently utilize and manage resources across multiple networks, leading to suboptimal performance and energy consumption.
Innovation Solution
A network of networks (NON) system that calculates and exposes capabilities of interconnected networks, allowing hosts to select the most suitable network based on their requirements and adjusts capabilities to match available resources, and includes a broker-less resource sharing mechanism for efficient resource utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If networks operate independently without resource sharing, then each network maintains its own resource management simplicity, but overall resource utilization is inefficient and energy consumption is suboptimal
Solution Approach 1:
The patent merges multiple independent networks into a unified network of networks (NON) where resources from different networks are pooled together. The resource pool aggregates computing, storage, and networking resources from multiple networks, enabling efficient allocation and sharing. This combining approach directly improves resource utilization and reduces energy consumption by eliminating redundant resources across networks.
Solution Approach 2:
The patent creates universal resource pools that can serve multiple networks and applications simultaneously. The resource pool is designed to be multi-functional, supporting various network types (5G, WiFi, wired) and workloads (compute, storage, networking). This universality allows resources to be dynamically allocated based on demand, improving overall utilization and reducing energy waste from dedicated but underutilized resources.
2Ease of operation
If networks expose detailed internal capabilities to external hosts, then hosts can make informed network selection decisions, but network security and resource control are compromised
Solution Approach 1:
The patent introduces a broker-less intermediary mechanism where networks expose only their external capabilities through standardized interfaces without revealing internal implementation details. The resource pool and network exposure mechanisms act as intermediaries that translate internal network capabilities into external service descriptions, allowing hosts to make informed decisions while maintaining security boundaries and resource control.
Solution Approach 2:
The patent applies local quality by allowing each network to expose different levels of capability information based on its specific security requirements and resource characteristics. Networks can selectively expose capabilities that match their security policies while still providing sufficient information for hosts to make appropriate network selection decisions. This localized approach balances transparency with security.
3Speed
If hosts attach to networks without capability matching, then connection establishment is simple and fast, but application performance requirements are not met
Solution Approach 1:
The patent implements preliminary action by having networks pre-calculate and expose their capabilities before hosts need to connect. The capability exposure mechanism prepares network capability information in advance, allowing hosts to quickly match their application requirements with suitable networks without complex real-time negotiations. This preliminary preparation maintains fast connection establishment while ensuring performance requirements are met.
Solution Approach 2:
The patent incorporates feedback mechanisms where hosts provide their application requirements to the network selection process, and networks respond with their capability information. This feedback loop enables capability-matched network selection while maintaining efficient connection establishment. The system learns from past selections and optimizes future matching decisions based on application performance requirements.
Data Source
AI summary
Method comprisingreading, from each of plural networks, a first capability indication exposed by the respective network, wherein the first capability indication exposed by the respective network indicates a first capability the respective network is capable to provide to a first host if the first host attaches to the respective network;checking, for each of the plural networks, whether the first capability the respective network is capable to provide matches a required first capability of a first application of the first host;instructing that the first host is attached to a first one of the plural networks in response to checking that the first capability the first one of the plural networks is capable to provide matches the required first capability of the first application of the first host.