DSP Noise Filter for Virtual Machine Network Test Signal Reliability

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In large-scale cloud-based data centers, detecting consistent network test result signal problems is time-consuming and costly due to the manual, piecemeal process of reviewing hardware and software issues across numerous computing nodes and virtual machines, with intermittent failures being difficult to distinguish from chronic ones.

Innovation Solution

A system employing a Digital Signal Processing (DSP) noise filter, such as a 'three strikes' method, is applied to network test result signals to differentiate between chronic and intermittent failures, filtering out intermittent issues and prioritizing chronic ones for investigation, using a virtual machine data store and a virtual machine test platform to manage and analyze test results.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual individualized review of hardware and software issues is performed on each node and virtual machine, then investigation thoroughness is improved, but time consumption and cost increase substantially

Engineering Contradiction:

Improveinvestigation thoroughnessVSAvoidtime consumption

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent segments the large-scale data center infrastructure into manageable units (nodes and virtual machines) and applies automated analysis to each segment. The system divides the investigation process into automated signal analysis and manual review phases, allowing thorough examination of individual components without requiring manual review of the entire system.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary automated analysis system that processes test result signals between the infrastructure components and human investigators. This intermediary system filters, aggregates, and prioritizes data, providing refined information to investigators without replacing the need for thorough analysis of critical issues.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Productivity

If automated systems are implemented to detect anomalies across numerous nodes and virtual machines, then productivity and efficiency are improved, but difficulty in detecting intermittent failures increases

Engineering Contradiction:

Improveanomaly detection efficiencyVSAvoidintermittent failure detection

Core Design Contradiction:

ProductivityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent implements preliminary automated analysis of test result signals to identify patterns, trends, and anomalies before human investigation. The system pre-processes data by filtering noise, aggregating results across multiple nodes, and prioritizing suspicious patterns, making intermittent failures more detectable through systematic preliminary examination rather than random sampling.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent employs feedback mechanisms where test results from multiple nodes and virtual machines are continuously monitored and fed back into the analysis system. This allows the system to learn from patterns across the infrastructure and improve detection of intermittent failures over time by recognizing recurring anomalies that may indicate underlying issues.

Inventive Principle:

Principle #23Feedback

3Reliability

If comprehensive testing of all virtual machines is performed, then reliability of service applications is improved, but resource consumption and cost increase

Engineering Contradiction:

Improveservice application reliabilityVSAvoidresource consumption

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The patent applies partial action by performing automated analysis on all virtual machines but focusing detailed investigation only on cases where anomalies are detected. The system processes test signals from the entire infrastructure efficiently through automation, then applies resources intensively only to suspicious cases, achieving comprehensive monitoring without the resource cost of exhaustive manual testing of every component.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS10812363B2Digital signal processing noise filter to increase test signal reliability

Publication Date: 2020.10.20 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Some embodiments are associated with a data center cloud computing environment executing a plurality of virtual machines. A virtual machine data store may contain information about the plurality of virtual machines associated with the cloud computing environment. A virtual machine test platform may access the information in the virtual machine data store and periodically initiate a network test for each of the plurality of virtual machines based on the information in the virtual machine data store. The virtual machine test platform may then receive network test result signals from the virtual machines in response to the executed network tests and apply, for each virtual machine, a digital signal processing noise filter to successive network test result signals from that virtual machine. An output of the digital signal processing noise filters may then be transmitted from the virtual machine test platform.