Optimization of application workflow in mobile embedded devices

Patent metadata

Field	Value
Publication number	US-2016378550-A1
Application number	US-201514950934-A
Country	US
Kind code	A1
Filing date	Nov 24, 2015
Priority date	Jun 29, 2015
Publication date	Dec 29, 2016
Grant date	—

Abstract

An aspect includes optimizing an application workflow. The optimizing includes characterizing the application workflow by determining at least one baseline metric related to an operational control knob of an embedded system processor. The application workflow performs a real-time computational task encountered by at least one mobile embedded system of a wirelessly connected cluster of systems supported by a server system. The optimizing of the application workflow further includes performing an optimization operation on the at least one baseline metric of the application workflow while satisfying at least one runtime constraint. An annotated workflow that is the result of performing the optimization operation is output.

First claim

What is claimed is:

1 . A method for optimizing an application workflow, comprising:
characterizing, by a processor coupled to a memory, the application workflow by determining at least one baseline metric related to an operational control knob of an embedded system processor, the application workflow configured to perform a real-time computational task encountered by at least one mobile embedded system of a wirelessly connected cluster of systems supported by a server system; performing, by the processor, an optimization operation on the at least one baseline metric of the application workflow while satisfying at least one runtime constraint; and outputting, by the processor, an annotated workflow, the annotated workflow being a result of the performing of the optimization operation.

2 . The method of claim 1 , wherein the optimization operation comprises a static preparation of the application workflow, the static preparation comprising:
characterizing the at least one baseline metric of the application workflow as a function of time; and optimizing the at least one baseline metric within the at least one runtime constraint.

3 . The method of claim 2 , wherein the static preparation executes a single optimization heuristic.

4 . The method of claim 2 , wherein the static preparation executes a linear optimization heuristic.

5 . The method of claim 2 , wherein the static preparation executes a structured direct acyclic graph optimization.

6 . The method of claim 1 , wherein the optimization operation comprises a dynamic optimization of the application workflow, the dynamic optimization comprising:
emulating a deployment environment of the mobile embedded system(s); and optimizing the at least one baseline metric of the application workflow operating in the deployment environment within the at least one runtime constraint.

7 . The method of claim 6 , wherein the dynamic optimization executes a turbo-boost optimization, the turbo boost optimization comprising a boosting of voltage-frequency levels used for short intervals to meet a real-time deadline of the emulated deployment environment.

8 . The method of claim 6 , wherein the dynamic optimization comprises:
injecting per-application execution time deviations into the emulated deployment environment.

9 . The method of claim 1 , wherein the at least one runtime constraint is a performance-related constraint of a real-time execution deadline.

10 . The method of claim 1 , wherein the optimization operation comprises adjusting a processor setting of one of the mobile embedded systems by changing a value of an adaptive power performance control knob, the processor setting corresponding to a baseline performance, a baseline power, or a baseline resilience of the one of the mobile embedded systems.

11 . The method of claim 1 , further comprising:
re-tuning the annotated workflow to factor in simulated run-time uncertainties caused by the operational environment of said embedded system processor.

12 . The method of claim 1 , wherein the optimization operation comprises a guard mechanism to minimize a number of failures during the optimization operation,
wherein each failure represents an inability by the optimization operation to find an optimization solution within the at least one runtime constraint.

13 . The method of claim 1 , comprising:
synthesizing, by the processor, a plurality of individual applications targeted for per-application optimization controls into a synthesized workflow based on the at least one runtime constraint, a plurality of metrics, hardware resource constraints of the mobile embedded systems, and software redundancy options; and receiving, by the processor, a synthesized workflow as the application workflow.

Assignees

• IBM

Inventors

• Bertran Monfort Ramon
• Bose Pradip
• Buyuktosunoglu Alper
• Cher Chen-Yong
• Jacobson Hans M
• Song William J
• Swaminathan Karthik V
• Vega Augusto J
• Wang Liang

Classifications

• G06F1/3296

  by lowering the supply or operating voltage · CPC title

• G06F9/4843

  by program, e.g. task dispatcher, supervisor, operating system · CPC title

• G06F9/4887 Primary

  involving deadlines, e.g. rate based, periodic · CPC title

• G06F9/543

  User-generated data transfer, e.g. clipboards, dynamic data exchange [DDE], object linking and embedding [OLE] · CPC title

• G06F9/4893

  taking into account power or heat criteria (power management in computers in general G06F1/3203; thermal management in computers in general G06F1/206) · CPC title