Processor Configuration Optimization via Mixed Integer Linear Programming

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Determining an optimal set of processors for complex hardware systems, such as avionics networks, is challenging due to the complexity of computational requirements and the need for real-time processing, leading to inadequate resource allocation and increased development time.

Innovation Solution

Formulating system configuration constraints as an objective function and a set of linear inequalities, optimizing over a convex polytope using mixed integer linear programming to determine the optimal number and capability of processors, ensuring adequate resource allocation and meeting operational and functional requirements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If manual assessment and estimation methods are used to determine processor requirements, then engineers can select processors based on assumed margins, but the process is time-consuming and may result in inadequate resource allocation

Engineering Contradiction:

Improveresource allocation adequacyVSAvoiddevelopment time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent replaces manual mechanical assessment processes with an automated mathematical optimization system. The processor selection and allocation are determined through mixed-integer linear programming that automatically evaluates computational requirements, constraints, and margins to produce optimal processor configurations without time-consuming manual iteration

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system transforms the processor selection problem from a qualitative manual assessment into a quantitative optimization problem by defining computational requirements, constraints, and objective functions in terms of mathematical parameters. This allows automated calculation of optimal processor quantities and capabilities based on system requirements

Inventive Principle:

Principle #35Parameter changes

2Productivity

If new generations of faster processors are used to meet computational processing power needs, then integration and complexity of the system can be increased, but the system weight and cost increase

Engineering Contradiction:

Improvecomputational processing powerVSAvoidsystem weight

Core Design Contradiction:

ProductivityVSWeight of moving object

Solution Approach 1:

The optimization system evaluates processor speed, quantity, and capability as adjustable parameters to find the minimum configuration that meets computational requirements. By mathematically determining the precise processor specifications needed, the system avoids over-provisioning with heavier, faster processors than necessary, thereby minimizing system weight while achieving required productivity

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

Instead of using excessive processor capacity to ensure meeting requirements, the optimization system calculates the precise partial amount of processing power needed. The mixed-integer linear programming determines the minimum processor quantity and speed that satisfies computational requirements, avoiding the weight penalty of excessive processing capacity

Inventive Principle:

Principle #16Partial or excessive action

3Reliability

If manual assessment of application needs for redundancy, separation, and I/O resources is performed, then engineers can estimate processor quantities, but the process is time-consuming and uncertain

Engineering Contradiction:

Improveprocessor selection confidenceVSAvoidsystem configuration complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent replaces manual engineering judgment and assessment with an automated mathematical optimization system. The mixed-integer linear programming framework automatically handles the complex evaluation of redundancy requirements, separation constraints, and I/O resource allocation, producing reliable processor selections without manual uncertainty

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system segments the complex processor selection problem into distinct mathematical components: computational requirements, redundancy constraints, separation constraints, and I/O resource constraints. Each segment is formulated as part of the overall optimization problem, allowing systematic and reliable solution of the complex configuration

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS8880381B2Optimization of processor characteristics and large scale system optimization through domain decomposition

Publication Date: 2014.11.04 THE BOEING CO

AI summary

Systems and methods for optimizing processor requirements for a complex hardware system are disclosed. A set of complex hardware system configuration constraints are formulated as an objective function and a set of linear inequalities, and a convex polytope is formed from the set of linear inequalities. The objective function is optimized over the convex polytope using mixed integer linear programming means to obtain an optimal solution. Processor requirements for the complex hardware system are determined based on the optimal solution.