Ad-hoc Network Node Grouping via Weight-Based Master Election

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for organizing nodes in large-scale ad-hoc networks are costly in terms of resources and limited in forming large groups, as they require each node to transmit weight information to many others, increasing resource consumption and system time.

Innovation Solution

A method where nodes calculate and store information about the most significant node with the highest weight, transmitting this information step-by-step, allowing nodes to determine a master node without needing to know all other nodes' weights, and repeating this process to form larger groups efficiently.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If each node transmits weight information to all other nodes in the network, then the master node can be accurately determined, but the resource consumption and system time increase significantly

Engineering Contradiction:
Improvemaster node determination accuracyVSAvoidresource consumption
Core Design Contradiction:
ReliabilityVSLoss of energy

Solution Approach 1:

The network is divided into clusters with master nodes, and the weight information propagation is segmented into multiple rounds. Each node only needs to transmit its weight information to a limited number of neighbors in each round, rather than broadcasting to all nodes simultaneously. This segmentation reduces the transmission overhead while ensuring that weight information eventually reaches all necessary nodes through iterative propagation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Nodes pre-calculate their weight values based on local parameters (mobility, transmission power, coverage area, battery power) before the master node election process. This preliminary calculation avoids the need for nodes to exchange and process complete weight information from all other nodes, as each node already has its own weight ready for comparison through limited local exchanges.

Inventive Principle:
Principle #10Preliminary action

2Loss of information

If each node transmits weight information to all other nodes, then complete network awareness is achieved, but the method is only suitable for small groups of nodes

Engineering Contradiction:
Improvenetwork information completenessVSAvoidgroup size scalability
Core Design Contradiction:
Loss of informationVSProductivity

Solution Approach 1:

The network is organized into clusters of nodes, each with a master node. This segmentation allows the network to scale to large sizes by dividing it into multiple manageable groups. Each cluster operates semi-independently, with master nodes handling inter-cluster communication, enabling the overall network to grow without requiring every node to communicate with every other node.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a hierarchical dimension to the network structure, organizing nodes into clusters with master nodes that operate at a higher level. This dimensional change from a flat all-to-all communication model to a hierarchical clustered model enables scalability to large networks while maintaining information completeness through the hierarchical information flow.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Reliability

If complete weight information is propagated across the network, then accurate master node selection is ensured, but the complexity of the organization method increases

Engineering Contradiction:
Improvemaster node selection accuracyVSAvoidorganization method complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

Nodes pre-calculate their weight values based on local parameters before the election process. This preliminary action simplifies the subsequent master node selection by eliminating the need for complex real-time information exchange and processing, as each node already has its weight determined and ready for comparison through simple local message passing.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The master node election process is structured as a periodic iterative procedure where weight information is propagated in discrete rounds. In each round, nodes exchange information with their neighbors and update their knowledge of the most significant node. This periodic structure simplifies the complexity by breaking down the continuous information exchange into manageable discrete steps, making the overall process more tractable and easier to implement.

Inventive Principle:
Principle #19Periodic action

Data Source

PatentEP1992109B1Method of organizing nodes of a network into groupings of nodes, computer program for implementing such a method and communication device forming a node of a network of nodes
Publication Date: 2016.05.04 ORANGE SA
  • EP1992109B1 patent drawingFigure 1
  • EP1992109B1 patent drawingFigure 2~4

AI summary

The method comprises a step (130) of calculating, for each node, a weight representing the suitability of this node to be the master node of a grouping, and a step (190) of electing a master node of a grouping, such that the weight of this master node is greater than or equal to the weights of the other nodes of the grouping. The method also comprises a step (140, 180) of storing, in each node, an item of information (14) containing an identifier of the node of highest weight known by this node and the value of this highest weight, a step (150) of transmitting, by each node to nodes adjacent to this node, the information item (14) stored in this node, and a step (170) of comparing, for each node, information items (14) transmitted to this node during the transmitting step (150) with the item of information stored in this node, so as to determine a new node of highest weight known by this node. The steps of storing (180), transmitting (150) and comparing (170) are repeated successively a predetermined number (k) of times, on completion of which the step (190) of electing a master node is carried out, in the course of which a node (M) whose stored item of information contains its own identifier is chosen to be a master node.