Granted Patent B2

US 12,463,919 · App. 18/509,873 · Granted Nov 4, 2025

Inventors: Anthony Rowe (Pittsburgh, PA); Carlee Joe-Wong (Pittsburgh, PA); Michael Pressler (Karlsruhe, DE); Nuno Pereira (Pittsburgh, PA); Tianshu Huang (Pittsburgh, PA)

Assignees: ROBERT BOSCH GMBH; CARNEGIE MELLON UNIVERSITY

H04L47/83H04L41/16H04L47/805

Abstract

Claims (27)

1 . A method for predicting resource usage for applications in a distributed system, comprising the following steps:

obtaining resource usage data, the resource usage data resulting from measuring resource usage of different applications on different devices of the distributed system;

detecting, by an orchestrator, a change and/or an event in the distributed system that requires a re-configuration of the distributed system;

predicting, by the orchestrator, the resource usage of at least one application when deployed on one or different devices of the distributed system, the predicting being carried out based on the obtained resource usage data;

and initiating the required re-configuration based on the detecting and the predicting, wherein the predicting of the resource usage is carried out based on matrix factorization, wherein features for the applications and devices are obtained and combined with features learned by the matrix factorization for the predicting of the resource usage.

2 . The method of claim 1 , wherein the re-configuration includes a deployment and/or mapping of the at least one application, the deployment and/or mapping being carried out based on the predicted resource usage such that the resource usage is optimized and/or requirements including real time and/or quality of service requirements for the at least one application are fulfilled.

3 . The method of claim 1 , wherein the detected change and/or event includes at least one of the following:

introduction of the at least one application as a new application of the distributed system,

introduction of a new hardware,

changes of real-time requirements for the applications,

new deployments of applications,

a change of the configuration of the distributed system,

a changed execution mode of the applications.

4 . The method of claim 1 , wherein the resource usage data resulting from a monitoring of multiple applications and/or device data resulting from a monitoring of multiple devices is combined, the combined data being used for the predicting the resource usage of the at least one application.

5 . The method of claim 1 , wherein the obtained resource usage data resulting from different applications and/or different devices is combined for being used by the predicting the resource usage, using matrix completion techniques.

6 . The method of claim 5 , wherein the predicting of the resource usage includes applying a machine learning model with the combined data.

7 . The method of claim 1 , wherein the applications are provided as bytecode, a number of times a bytecode instruction is executed being measured and used as a feature obtained by dynamic analysis for each application, and hardware descriptions for each device being recorded and additionally used as a feature, the features being combined with matrix factorization to learn feature embeddings for each application and device to predict the resource usage.

8 . A non-transitory computer-readable medium on which is stored a computer program including instructions for predicting resource usage for applications in a distributed system, the instructions, when executed by a computer, causing the computer to perform the following steps:

obtaining resource usage data, the resource usage data resulting from measuring resource usage of different applications on different devices of the distributed system;

detecting, by an orchestrator, a change and/or an event in the distributed system that requires a re-configuration of the distributed system;

predicting, by the orchestrator, the resource usage of at least one application when deployed on one or different devices of the distributed system, the predicting being carried out based on the obtained resource usage data;

and initiating the required re-configuration based on the detecting and the predicting, wherein the predicting of the resource usage is carried out based on matrix factorization, wherein features for the applications and devices are obtained and combined with features learned by the matrix factorization for the predicting of the resource usage.

9 . A data processing apparatus configured to predict resource usage for applications in a distributed system, the data processing apparatus configured to:

obtain resource usage data, the resource usage data resulting from measuring resource usage of different applications on different devices of the distributed system;

detect, by an orchestrator, a change and/or an event in the distributed system that requires a re-configuration of the distributed system;

predict, by the orchestrator, the resource usage of at least one application when deployed on one or different devices of the distributed system, the predicting being carried out based on the obtained resource usage data;

and initiate the required re-configuration based on the detecting and the predicting, wherein the predicting of the resource usage is carried out based on matrix factorization, wherein features for the applications and devices are obtained and combined with features learned by the matrix factorization for the predicting of the resource usage.

Assignments (2)

Priority Claims (1)

Continuity (1)

References Cited (18)