Dynamic ICN Routing for Compute Resource Selection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current Information Centric Network (ICN) forwarding strategies are not optimized for dynamic computation, as they typically rely on single compute servers and do not efficiently route computation requests to the most suitable compute resources, leading to suboptimal performance and increased overhead.

Innovation Solution

Implement a dynamic forwarding strategy that utilizes a new ContentType for data packets to inform intermediary nodes of compute resource availability, allowing routers to dynamically select the best compute server based on performance metrics and resource capabilities, thereby optimizing the routing of computation interest packets.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If traditional ICN forwarding strategies are used, then the network can route computation requests, but the routing is not optimized and does not consider compute resource availability and performance metrics

Engineering Contradiction:

Improverouting optimizationVSAvoidforwarding strategy complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by having compute nodes proactively publish their availability status and performance metrics to the network before receiving computation requests. This allows intermediary nodes to have routing decisions based on pre-collected information about compute resource capabilities, rather than making routing decisions without such information.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent implements feedback mechanisms where compute nodes continuously update their availability status and performance metrics in the network. Intermediary nodes use this feedback information to dynamically adjust routing decisions, selecting compute servers based on current network conditions and resource availability rather than static routing tables.

Inventive Principle:

Principle #23Feedback

2Productivity

If single compute servers are used, then the forwarding strategy is simple, but the performance is suboptimal and overhead is increased

Engineering Contradiction:

Improvecomputation processing efficiencyVSAvoidcompute resource management

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent applies dynamics by enabling the network to dynamically select compute servers based on real-time availability status and performance metrics. Instead of relying on a single static compute server, the system can adaptively route computation requests to the most suitable compute resource available at any given moment, improving overall system productivity.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent implements universality by creating a multi-functional compute resource selection mechanism that can handle various computation request types and route them to appropriate compute servers based on capabilities. The system universally supports multiple compute nodes with different resources, allowing flexible allocation across diverse workloads.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Reliability

If compute resource information is not shared in the network, then the forwarding strategy is simple, but intelligent routing to suitable compute resources cannot be achieved

Engineering Contradiction:

Improvecompute resource selection accuracyVSAvoidinformation sharing mechanism

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent uses intermediary nodes as mediators that collect, store, and forward compute resource information throughout the network. These intermediaries facilitate the sharing of availability status and performance metrics between compute nodes and routing decisions, enabling reliable compute resource selection without requiring direct peer-to-peer communication between all nodes.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS11659066B2Dynamic computation in an information centric network

Publication Date: 2023.05.23 INTEL CORP

AI summary

Systems and techniques for dynamic computation in an information centric network (ICN) are described herein. An interest packet to perform a computation may be received at a first interface of an ICN node. The ICN node may then perform a lookup in a forwarding information base (FIB) to identify a second interface to forward the interest packet. The interest packet may be forwarded on the second interface. Upon receipt of a data packet on the second interface in response to the interest packet, the ICN node may update an entry in the FIB for the second interfaces with a processing payload included in the data packet. The ICN node may then transmit the data packet downstream towards the originator of the interest packet.