Granted Patent B2

US 12,549,445 · App. 18/499,844 · Granted Feb 10, 2026

Inventors: David Burth Kurka (Campinas, BR); Diego Vrague Noble (Pelotas, BR); Karen Braga Enes (Belo Horizonte, BR); Thais Luca Marques De Almeida (São Gonçalo, BR); Italo Gomes Santana (Rio de Janeiro, BR)

Assignee: Dell Products L.P.

H04L41/12H04L41/0836H04L41/0894

Abstract

Claims (51)

1 . A system comprising:

at least one processing device including a processor coupled to a memory;

the at least one processing device being configured to implement the following steps:

using a community detection algorithm and network centrality metrics to identify vulnerable communities among edge nodes and inter-community connections between communities detected in an edge network, the edge nodes being grouped into a plurality of communities in the edge network, the community detection algorithm including a Louvain algorithm or a Leiden algorithm;

using a count of inter-community connections for the communities and a measure of network robustness to rank the communities according to a likelihood of being disconnected due to a network failure;

for each ranked community, using the detected inter-community connections to determine a connection policy to create new inter-community connections to neighboring communities;

generating a list of connection policies ranked by importance of inter-community connections;

applying at least a subset of the connection policies to the edge network according to the generated list;

using the community detection algorithm to identify communities having low connectivity:

identifying inter-community edges that are overloaded;

determining a connection policy to create the new inter-community connections in response to a trigger event;

determining that the new inter-community connections are sufficient to reconnect the identified communities within a predetermined threshold, the threshold being a ratio between a count of intra-community connections and a count of inter-community connections; and

in response to determining the new inter-community connections are sufficient, applying the connection policy to the edge network.

2 . The system of claim 1 , wherein the inter-community edges are overloaded due to a flow increase.

3 . The system of claim 1 , wherein the threshold is determined based on the new inter-community connections.

4 . The system of claim 1 , wherein the threshold is a minimum count of inter-community connections.

5 . The system of claim 1 , wherein the steps are performed in response to the trigger event.

6 . The system of claim 1 , wherein the trigger event is a predetermined number of edge devices joining or leaving the edge network.

7 . The system of claim 1 , wherein the at least one processing device is further configured to implement the following steps:

using the community detection algorithm to classify edges between edge nodes as inter-community edges or intra-community edges; and

ranking the edges according to a connectivity importance measure.

8 . The system of claim 7 , wherein the connectivity importance measure is a betweenness centrality measure for the edges.

9 . The system of claim 1 , wherein the connection policy is configured to create the new inter-community connections by using a structure of the communities to add or remove connections among the communities to increase a robustness of the edge network.

10 . The system of claim 1 , wherein the steps are performed upon a determination that a count of changes in a map of the edge network exceeds a predetermined threshold compared with a previous map of the edge network.

11 . The system of claim 1 , wherein the communities are represented using undirected graphs.

12 . The system of claim 1 , wherein the communities are densely connected.

13 . The system of claim 1 , wherein the communities are assortative.

14 . The system of claim 1 , wherein the edge nodes are edge devices.

15 . A method comprising:

using a community detection algorithm and network centrality metrics to identify vulnerable communities among edge nodes and inter-community connections between communities detected in an edge network, the edge nodes being grouped into a plurality of communities in the edge network, the community detection algorithm including a Louvain algorithm or a Leiden algorithm;

using a count of inter-community connections for the communities and a measure of network robustness to rank the communities according to a likelihood of being disconnected due to a network failure;

for each ranked community, using the detected inter-community connections to determine a connection policy to create new inter-community connections to neighboring communities;

generating a list of connection policies ranked by importance of inter-community connections;

applying at least a subset of the connection policies to the edge network according to the generated list;

using the community detection algorithm to identify communities having low connectivity;

identifying inter-community edges that are overloaded;

determining a connection policy to create the new inter-community connections in response to a trigger event;

determining that the new inter-community connections are sufficient to reconnect the identified communities within a predetermined threshold, the threshold being a ratio between a count of intra-community connections and a count of inter-community connections; and

in response to determining the new inter-community connections are sufficient, applying the connection policy to the edge network.

16 . The method of claim 15 , wherein the likelihood of being disconnected due to the network failure is determined using a vulnerability score for each community, and wherein the vulnerability score is determined based on the count of inter-community connections and on the measure of network robustness.

17 . A non-transitory processor-readable storage medium having stored thereon program code of one or more software programs, wherein the program code when executed by at least one processing device causes the at least one processing device to perform the following steps:

using a community detection algorithm and network centrality metrics to identify vulnerable communities among edge nodes and inter-community connections between communities detected in an edge network, the edge nodes being grouped into a plurality of communities in the edge network, the community detection algorithm including a Louvain algorithm or a Leiden algorithm;

using a count of inter-community connections for the communities and a measure of network robustness to rank the communities according to a likelihood of being disconnected due to a network failure;

for each ranked community, using the detected inter-community connections to determine a connection policy to create new inter-community connections to neighboring communities;

generating a list of connection policies ranked by importance of inter-community connections;

applying at least a subset of the connection policies to the edge network according to the generated list;

using the community detection algorithm to identify communities having low connectivity;

identifying inter-community edges that are overloaded;

determining a connection policy to create the new inter-community connections in response to a trigger event;

determining that the new inter-community connections are sufficient to reconnect the identified communities within a predetermined threshold, the threshold being a ratio between a count of intra-community connections and a count of inter-community connections; and

in response to determining the new inter-community connections are sufficient, applying the connection policy to the edge network.

Assignments (1)

Continuity (1)

References Cited (11)