Intermittent Network Routing Using Inter-Contact Time Matrices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing routing methods for intermittently connected networks, such as Delay Tolerant Networks, face challenges in minimizing delivery times due to high loss rates and long propagation delays, leading to inefficient message routing and low delivery rates.
Innovation Solution
A multi-hop routing method that estimates average inter-contact times between nodes and determines optimal relay paths to minimize delivery times, using a matrix to calculate hop delays and select relay nodes, ensuring loop-free routing and efficient message transfer.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional IP routing mechanisms are used in intermittently connected networks, then routing simplicity is maintained, but message delivery fails due to high loss rates and long propagation delays
Solution Approach 1:
The patent introduces a store-and-forward mechanism where intermediate nodes act as mediators to hold messages until the destination becomes available. This resolves the contradiction by enabling reliable delivery in intermittent networks without requiring complex end-to-end path management, as each node independently manages its own message forwarding decisions based on encounter history.
Solution Approach 2:
The patent implements preliminary action by pre-calculating and storing encounter history matrices and inter-contact time statistics before routing decisions are needed. This allows nodes to make informed routing choices when encounters occur, improving delivery reliability without adding complexity during actual message transmission.
2Reliability
If probabilistic routing methods are used to handle intermittent connectivity, then message delivery is achieved, but delivery time estimation becomes difficult due to lack of inter-contact time consideration
Solution Approach 1:
The patent implements feedback by continuously updating the encounter history matrix with actual inter-contact time measurements. This feedback loop enables accurate delivery time estimation while maintaining high delivery rates, as the system learns from past encounters to predict future availability and optimize routing decisions accordingly.
Solution Approach 2:
The patent performs preliminary calculation of expected delivery times using the encounter history matrix before actual message transmission. This allows the system to provide accurate delivery time estimates to applications while maintaining reliable message delivery through store-and-forward mechanisms.
3Loss of time
If multi-hop routing with optimal relay selection is implemented, then delivery time is minimized, but routing calculation complexity increases
Solution Approach 1:
The patent implements self-service by enabling each node to independently calculate its own optimal relay nodes using the encounter history matrix. This decentralized approach minimizes delivery time through optimal relay selection while avoiding the complexity of centralized routing calculations, as each node autonomously makes routing decisions based on locally available historical data.
Solution Approach 2:
The patent applies partial action by considering only the top-k relay nodes with the highest encounter frequencies rather than evaluating all possible paths. This reduces calculation complexity significantly while still achieving near-optimal delivery times, as the most frequently encountered nodes provide the best forwarding opportunities.
4Ease of operation
If strict shortest path routing is enforced, then routing path is well-defined, but opportunistic transmission possibilities are limited and delivery efficiency decreases
Solution Approach 1:
The patent implements dynamics by allowing routing paths to adapt dynamically based on real-time encounter opportunities rather than following fixed predetermined paths. Nodes can deviate from the strict shortest path when encountering better forwarding opportunities, improving delivery efficiency while maintaining sufficient path definition through the encounter history-based relay selection.
Data Source
Figure 1~5
Figure 2
AI summary
The invention relates to a routing method intended for intermittently connected networks. The inventive method is characterised in that, in order to determine routing information for an opportunistically defined routing, it includes at least the following steps: for a network comprising a number n of nodes, wherein kmax is the maximum number of relays to be considered in the routing and M1 is a matrix of size nxn, initialisation M1 = ij> of the estimations of the average inter-contact times for nodes i and j. If k+1 corresponds to the number of relays considered for the routing, k=0 is initialised. And at step k, for each pair of nodes (i, j) (with 1<=i<=n, 1 <=j<=n, i?j), estimation of the delay at (k+1) jumps and the list of corresponding relays from ij> matrix of the inverses of the average inter-contact times and estimations of the delays at k jumps present in Mkand storing of the estimated inter-contact delays in matrix M(k+1)(i,j) and the list of corresponding relays Vi(d)k+1, (M(k+1 )(i,j), Vi(d)k+1), where Vj(d)k+1corresponds to the list of relay nodes considered at node i when a message must be transmitted to d for a number of relays k+1. If k+1 = kmax, the steps are stopped, otherwise k=k+1 is implemented.