Measureless Signal Decomposition for Data Server Diagnostics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Classical monitoring methods for complex systems like data servers fail to effectively leverage statistical distribution properties, leading to high rates of false or missed alarms due to variables with infinite moments, which lack meaningful mean and variance estimates.
Innovation Solution
A method involving a monitoring facility with a central processing unit that generates probing characteristics, decomposes measureless signals into component parts using empirical mode decomposition, reconstructs signals, and processes them to exclude offending components, thereby generating a well-behaved approximation for accurate diagnostic information extraction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If classical monitoring methods are used to monitor data server metrics, then system monitoring coverage is achieved, but false alarms and missed alarms occur at high rates due to variables with infinite moments
Solution Approach 1:
The patent transforms the monitoring approach by changing the parameters used for analysis. Instead of directly analyzing raw metrics with infinite moments, the system applies mathematical transformations (characteristic functions, Fourier transforms) to convert the data into a form where statistical properties can be meaningfully extracted. This parameter transformation resolves the contradiction by enabling reliable statistical analysis without losing the original data's informational content.
Solution Approach 2:
The patent introduces characteristic functions and Fourier transforms as intermediary mathematical tools between the raw metrics and the statistical analysis. These intermediaries allow the system to handle variables with infinite moments by transforming them into a domain where meaningful statistical properties emerge, thereby improving alarm accuracy while maintaining measurement precision.
2Reliability
If monitoring is performed on a large number of various metrics, then comprehensive system coverage is achieved, but detecting abnormalities becomes onerous and time-consuming
Solution Approach 1:
The patent creates a universal monitoring framework that can handle multiple different metric types through a common mathematical approach. The characteristic function and Fourier transform methodology serves as a multi-functional tool that works across diverse metric types, enabling comprehensive system coverage while streamlining the detection process through a unified analysis method.
Solution Approach 2:
By transforming multiple different metrics into a common transformed domain using Fourier transforms, the system enables parallel and efficient analysis of all metrics. This parameter transformation approach allows comprehensive monitoring coverage while reducing the time required to detect abnormalities across all system metrics simultaneously.
3Quantity of substance
If variables with infinite moments are analyzed using classical statistical methods, then all available data is utilized, but meaningful mean and variance estimates cannot be obtained
Solution Approach 1:
The patent introduces characteristic functions as an intermediary mathematical construct that bridges the gap between raw data with infinite moments and meaningful statistical estimates. The characteristic function captures all information from the original data while existing in a transformed domain where moments can be derived, thus utilizing all available data while obtaining precise mean and variance estimates.
Solution Approach 2:
The patent replaces classical mechanical statistical methods (direct calculation of mean and variance) with a transformed domain approach using Fourier transforms and characteristic functions. This substitution allows the system to work with variables that have infinite moments by operating in a mathematical domain where these moments become well-defined and computable.
Data Source
AI summary
In general the invention relates to a method for processing signals from a data server. The method includes obtaining, by a monitoring facility, a measured signal from the data server, wherein the monitoring facility comprises a central processing unit, generating a first probing characteristic from the measured signal, and determining that the first probing characteristic is measureless. The method further includes decomposing, by the central processing unit in response to the determination, the first probing characteristic into a plurality of component parts, constructing a reconstructed signal using a first one of the plurality of component parts, generating a second probing characteristic using the reconstructed signal, wherein the second probing characteristic is not measureless, and processing the reconstructed signal by the monitoring facility.