Intelligent resource allocation based on security profile of edge device network

A method, comprising:

• determining properties of an edge device network, the edge device network comprising a plurality of edge devices configured to execute computing tasks of the edge network;
• transforming the determined properties into device changes data, wherein the device changes data indicates a potential malicious attack on a first edge device of the plurality of edge devices based on an increase in an amount of memory utilized within a time interval by the first edge device;
• determining an anomaly is detected based at least in part on the device changes data of the edge device network, wherein the device changes data comprise features of the edge device network that are linked to a potential security compromise or that impact efficiency of the edge device network, and further wherein the anomaly is determined by determining that one or more properties of the edge device network are outside a predefined range of property values;
• after determining the anomaly is detected, executing a first artificial intelligence (AI) algorithm using real attack data and input noise data as input to train the first AI algorithm, wherein the first AI algorithm is configured to generate as output one or more generated attacks, wherein the one or more generated attacks are utilized by a quantum computing subsystem as one or more simulated attacks for testing a resiliency of the edge device network based on the device changes data and wherein the first AI algorithm is a deconvolutional neural network;
• training a second AI algorithm with the one or more generated attacks and the real attack data as input to generate the device changes data as output, wherein the second AI algorithm is a convolutional neural network;
• determining, by the second AI algorithm, an error during the training of the second AI algorithm;
• inputting the determined error into the first AI algorithm to update the first AI algorithm;
• executing at least one of the one or more simulated attacks in a plurality of available configurations of the edge device network;
• determining, based at least in part on results of the executed simulated attacks, a security score for each of the plurality of configurations;
• determining a selected configuration of the plurality of available configurations of the edge device network based at least in part on the security score determined for each of the plurality of configurations;
• and
• implementing the selected configuration at the edge device network by allocating additional computational resources to the first edge device that was utilizing the increased amount of memory.

An edge network management system comprising a resource deployment subsystem and a quantum computing subsystem, wherein:

• the resource deployment subsystem comprises a processor configured to:
• determine properties of an edge device network, the edge device network comprising a plurality of edge devices configured to execute computing tasks of the edge network;
• transform the determined properties into device changes data, wherein the device changes data indicates a potential malicious attack on a first edge device of the plurality of edge devices based on an increase in an amount of memory utilized within a time interval by the first edge device;
• determine an anomaly is detected based at least in part on the device changes data of the edge device network, wherein the device changes data comprise features of the edge device network that are linked to a potential security compromise or that impact efficiency of the edge device network, and further wherein the anomaly is determined by determining that one or more properties of the edge device network are outside a predefined range of property values;
• after determining the anomaly is detected, execute a first artificial intelligence (AI) algorithm using real attack data and input noise data as input to train the first AI algorithm, wherein the first AI algorithm is configured to generate as output one or more generated attacks, wherein the one or more generated attacks are utilized by the quantum computing subsystem as one or more simulated attacks for testing a resiliency of the edge device network based on the device changes data and wherein the first AI algorithm is a deconvolutional neural network;
• train a second AI algorithm with the one or more generated attacks and the real attack data as input to generate the device changes data as output, wherein the second AI algorithm is a convolutional neural network;
• determine, by the second AI algorithm, an error during the training of the second AI algorithm;
• and
• input the determined error into the first AI algorithm to update the first AI algorithm;
• the quantum computing subsystem is configured to:
• execute at least one of the one or more simulated attacks in a plurality of available configurations of the edge device network;
• and
• determine, based at least in part on results of the executed simulated attacks, a security score for each of the plurality of configurations;
• and
• the resource deployment subsystem is further configured to:
• determine a selected configuration of the plurality of available configurations of the edge device network based at least in part on the security score determined for each of the plurality of configurations;
• and
• implement the selected configuration at the edge device network by allocating additional computational resources to the first edge device that was utilizing the increased amount of memory.