Self-healing and resiliency in radio-based networks using a community model
Disclosed are various embodiments for improving self-healing and resiliency of radio-based networks through the use of a community model. In one embodiment, a community model graph representing a radio-based network is generated. Individual nodes of the community model graph correspond to one or more network functions of the radio-based network. Individual edges of the community model graph correspond to a respective network connection between the network function(s) of the radio-based network. The individual edges are assigned a respective measure of relative importance. An adverse health event is detected that impacts a plurality of network connections between network functions in the radio-based network. Self-healing activities for the network connections are prioritized based at least in part on the respective measures of relative importance corresponding to individual network connections.
1. A system, comprising:
a radio-based network at least partly implemented in a cloud provider network; and
at least one computing device in the cloud provider network configured to at least:
generate a community model graph representing the radio-based network, wherein individual nodes of the community model graph are assigned to logical communities encompassing one or more network functions of the radio-based network, individual edges of the community model graph correspond to a respective network connection between the one or more network functions of the radio-based network, and the individual edges are assigned a respective measure of relative importance;
select a particular network connection between the one or more network functions having a highest measure of relative importance of the particular network connection; and
initiate at least one of: a bandwidth upgrade for the particular network connection or a deployment of an additional network connection to supplement the particular network connection.
2. The system of claim 1 , wherein the particular network connection connects a radio access network (RAN)-enabled edge server with the cloud provider network.
3. The system of claim 1 , wherein the community model graph is constructed using a Girvan-Newman algorithm.
4. The system of claim 1 , wherein the at least one computing device is further configured to at least modify a service function chain for a service function forwarder to route network traffic from a source to a first network function instead of a second network function associated with the particular network connection.
5. The system of claim 1 , wherein the at least one computing device is further configured to at least:
detect an adverse health event impacting a plurality of network connections between network functions in the radio-based network; and
prioritize self-healing activities for the plurality of network connections based at least in part on the respective measures of relative importance corresponding to individual ones of the plurality of network connections.
6. The system of claim 1 , wherein the respective measure of relative importance corresponds to an edge betweenness value.
7. A computer-implemented method, comprising:
generating a community model graph representing a radio-based network, wherein individual nodes of the community model graph correspond to one or more network functions of the radio-based network, individual edges of the community model graph correspond to a respective network connection between the one or more network functions of the radio-based network, and the individual edges are assigned a respective measure of relative importance;
selecting a particular network connection between the one or more network functions based at least in part on the respective measure of relative importance of the particular network connection; and
implementing an action with respect to the particular network connection to increase resiliency of the radio-based network.
8. The computer-implemented method of claim 7 , wherein the community model graph is generated based at least in part on a Girvan-Newman algorithm.
9. The computer-implemented method of claim 7 , wherein the respective measure of relative importance corresponds to an edge betweenness value.
10. The computer-implemented method of claim 7 , wherein the action comprises initiating a bandwidth upgrade for the particular network connection.
11. The computer-implemented method of claim 7 , wherein the action comprises initiating a deployment of an additional network connection to supplement the particular network connection.
12. The computer-implemented method of claim 7 , wherein the action comprises configuring a service function forwarder to route network traffic from a source to a different network function associated with a different network connection based at least in part on modifying a service function chain.
13. The computer-implemented method of claim 7 , wherein the particular network connection connects a radio access network (RAN)-enabled edge server with a cloud provider network.
14. The computer-implemented method of claim 7 , wherein the one or more network functions comprise one or more core network functions for the radio-based network.
15. A computer-implemented method, comprising:
generating a community model graph representing a radio-based network, wherein individual nodes of the community model graph correspond to one or more network functions of the radio-based network, individual edges of the community model graph correspond to a respective network connection between the one or more network functions of the radio-based network, and the individual edges are assigned a respective measure of relative importance;
detecting an adverse health event impacting a plurality of network connections between network functions in the radio-based network; and
prioritizing self-healing activities for the plurality of network connections based at least in part on the respective measures of relative importance corresponding to individual ones of the plurality of network connections.
16. The computer-implemented method of claim 15 , wherein the adverse health event comprises at least one of: a failure of networking hardware or a failure of a physical data link.
17. The computer-implemented method of claim 15 , wherein the adverse health event comprises an overutilization of a networking component shared at least in part by the plurality of network connections.
18. The computer-implemented method of claim 15 , wherein prioritizing the self-healing activities further comprises:
determining that a first network connection of the plurality of network connections is associated with a higher measure of relative importance than a second network connection of the plurality of network connections; and
initiating a first remedial action to address the adverse health event with respect to the first network connection before initiating a second remedial action to address the adverse health event with respect to the second network connection.
19. The computer-implemented method of claim 15 , further comprising:
identifying a different network connection that is not impacted by the adverse health event based at least in part on the respective measure of relative importance of the different network connection; and
wherein the self-healing activities comprise redirecting network traffic from a particular network function to the different network connection.
20. The computer-implemented method of claim 15 , wherein the self-healing activities further comprise configuring a service function forwarder to route network traffic from a source to a different network function based at least in part on modifying a service function chain.