Adaptive resilient network communication

1 . A system for adaptive resilient network communication, the system comprising:

processing circuitry; and

memory, coupled to the processing circuitry, the memory including instructions stored thereon that, when executed by the processing circuitry cause the processing circuitry to perform operations, the operations comprising:

monitoring network traffic on multiple pathways between a user equipment and an application or a service at a network destination, wherein the multiple pathways include pathways of different network types, and wherein the different network types include at least two of: a broadband network, a wireless network, or a mobile network;

gathering network telemetry data from the monitored network traffic;

inputting the network telemetry data into a trained artificial intelligence model, wherein the trained artificial intelligence model receives additional data from an application programming interface (API) via an end-to-end socket, and wherein the additional data includes at least one of a Quality of Service (QOS) requirement, a reliability requirement, or a latency requirement of the application;

classifying the network telemetry data using the trained artificial intelligence model;

determining, using the trained artificial intelligence model, an anomaly condition in at least a portion of the multiple pathways; and

in response to the determining, selecting a mitigation technique for the at least a portion of the multiple pathways.

2 . The system of claim 1 , wherein the anomaly condition corresponds to a current condition in the at least a portion of the multiple pathways.

3 . The system of claim 1 , wherein the anomaly condition corresponds to a predicted future condition in the at least a portion of the multiple pathways, and wherein the predicted future condition is based at least in part on a real-time location of the user equipment.

4 . The system of claim 1 , wherein the at least a portion of the multiple pathways includes at least one of:

a radio access point, an access network, a core network, a network endpoint, or a path therebetween in one or more network types.

5 . The system of claim 1 , the operations further comprising:

training the trained artificial intelligence model using the classified network telemetry data.

6 . The system of claim 1 , wherein the network telemetry data includes at least one of: a channel quality, a real-time network load, a sub-network condition, or a node radiality.

7 . The system of claim 1 , wherein the user equipment is connected to a particular active pathway of the multiple pathways, wherein the particular active pathway is of a first type, and wherein the mitigation technique includes:

at least one of activating or adding a second particular pathway of the multiple pathways, or replicating at least a portion of the particular active pathway; and

routing at least a portion of data or traffic from the user equipment on at least one of the second particular pathway or the replicated portion of the particular active pathway or re-routing at least a portion of data or traffic from the user equipment to a different active pathway type.

8 . The system of claim 1 , wherein the multiple pathways further include a pathway on a satellite network.

9 . The system of claim 1 , the operations further comprising:

determining a risk score for the network telemetry data over one or more of the multiple pathways; and

comparing the risk score to a lower thresholding value and an upper thresholding value, and wherein the mitigation technique is selected based at least in part on the risk score.

10 . The system of claim 9 , wherein in response to the risk score falling between the lower thresholding value and the upper thresholding value, the risk score is recalculated until a moving window average for the risk score and a latest value of the risk score drops below the lower thresholding value or crosses above the upper thresholding value.

11 . A method for implementing adaptive resilient network communication, the method comprising:

monitoring network traffic on multiple pathways between a user equipment and an application or a micro-service at a network destination, wherein the multiple pathways include pathways of different network types, and wherein the different network types include at least two of:

a broadband network, a wireless network, or a mobile network;

gathering network telemetry data from the monitored network traffic;

inputting the network telemetry data into a trained artificial intelligence model;

receiving, by the trained artificial intelligence model, additional data from an application programming interface (API) via an end-to-end socket, wherein the additional data includes at least one of a Quality of Service (QOS) requirement, a reliability requirement, or a latency requirement of the application;

classifying the network telemetry data using the trained artificial intelligence model;

determining, using the trained artificial intelligence model, an anomaly condition in at least a portion of the multiple pathways; and

in response to the determining, selecting a mitigation technique for at least one pathway of the multiple pathways.

12 . The method of claim 11 , wherein the anomaly condition corresponds to a current condition in the at least a portion of the multiple pathways.

13 . The method of claim 11 , wherein the anomaly condition corresponds to a predicted future condition in the at least a portion of the multiple pathways, and wherein the predicted future condition is based at least in part on a real-time location of the user equipment.

14 . The method of claim 11 , wherein the user equipment is connected to a particular active pathway of the multiple pathways and the mitigation technique includes:

at least one of activating or adding a second particular pathway of the multiple pathways, or replicating at least a portion of the particular active pathway; and

routing at least a portion of data or traffic from the user equipment on at least one of the second particular pathway or the replicated portion of the particular active pathway.

15 . The method of claim 11 , further comprising:

determining a risk score for the network telemetry data over one or more of the multiple pathways; and

comparing the risk score to a lower thresholding value and an upper thresholding value, and wherein the mitigation technique is selected based at least in part on the risk score.

16 . The method of claim 15 , wherein in response to the risk score falling between the lower thresholding value and the upper thresholding value, the risk score is recalculated until a moving window average for the risk score and a latest value of the risk score drops below the lower thresholding value or crosses above the upper thresholding value.

17 . The method of claim 11 , wherein the mitigation technique includes:

replicating or adding at least one of a network service, a network resource, or a network component; and

migrating at least one of a task, at least a portion of the monitored network traffic, or data from the user equipment to the at least one of the replicated or added network service, network resource, or network component.

18 . At least one non-transitory machine-readable medium with instructions stored thereon, which when executed by a processor of a computing device, cause the processor to execute operations, the operations comprising:

monitoring network traffic on multiple pathways between a user equipment and an application or a service at a network destination, wherein the multiple pathways include pathways of different network types, and wherein the different network types include at least two of: a broadband network, a wireless network, or a mobile network;

gathering network telemetry data from the monitored network traffic;

inputting the network telemetry data into a trained artificial intelligence model, wherein the trained artificial intelligence model receives additional data from an application programming interface (API) via an end-to-end socket, and wherein the additional data includes at least one of a Quality of Service (QOS) requirement, a reliability requirement, or a latency requirement of the application;

classifying the network telemetry data using the trained artificial intelligence model;

determining, using the trained artificial intelligence model, an anomaly condition in at least a portion of the multiple pathways; and

in response to the determining, selecting a mitigation technique for at least one pathway of the multiple pathways.

19 . The at least one non-transitory machine-readable medium of claim 18 , wherein the anomaly condition corresponds to at least one of a current condition in the at least a portion of the multiple pathways, or a predicted future condition in the at least a portion of the multiple pathways, and wherein the predicted future condition is based at least in part on a real-time location of the user equipment.

20 . The at least one non-transitory machine-readable medium of claim 18 , wherein the at least a portion of the multiple pathways includes at least one of: a radio access point, an access network, a core network, a network endpoint, or a path therebetween in the different network types.

21 . The at least one non-transitory machine-readable medium of claim 18 , the operations further comprising:

training the trained artificial intelligence model using the classified network telemetry data.

22 . The at least one non-transitory machine-readable medium of claim 18 , wherein the user equipment is connected to a particular active pathway of the multiple pathways and wherein the mitigation technique includes:

at least one of activating or adding a second particular pathway of the multiple pathways, or replicating at least a portion of the particular active pathway; and

routing at least a portion of data or traffic from the user equipment on at least one of the second particular pathway or the replicated portion of the particular active pathway.

23 . The at least one non-transitory machine-readable medium of claim 18 , the operations further comprising:

determining a risk score for the network telemetry data over one or more of the multiple pathways; and

comparing the risk score to a lower thresholding value and an upper thresholding value, and wherein the mitigation technique is selected based at least in part on the risk score;

wherein in response to the risk score falling between the lower thresholding value and the upper thresholding value, the risk score is recalculated until a moving window average for the risk score and a latest value of the risk score drops below the lower thresholding value or crosses above the upper thresholding value.