Telemetry Bandwidth Enhancement via Correlated Signal Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High sensor density in computer servers leads to bandwidth limitations in system buses, resulting in slow sampling rates of telemetry data, which impede dynamic monitoring and fault detection in enterprise computing systems.
Innovation Solution
A system that groups telemetry data into clusters of correlated signals, omits some signals during sampling using a round-robin technique, and estimates the omitted signals using nonlinear, nonparametric regression, specifically multivariate state estimation (MSET), to increase bandwidth and enhance monitoring efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If sensor density is increased to improve monitoring coverage, then monitoring completeness is improved, but bandwidth limitations cause sampling rate to decrease
Solution Approach 1:
The patent segments the set of correlated telemetric signals into groups where signals within each group are highly correlated. By sampling only one signal from each group at full rate and using regression to estimate others, the system effectively segments the bandwidth allocation across correlated signals, resolving the contradiction between monitoring all signals and maintaining sampling rate.
Solution Approach 2:
The patent changes the parameter of signal representation by transforming correlated signals into a reduced set of independent samples plus regression residuals. This parameter transformation allows the system to maintain effective monitoring of all signals while reducing the actual bandwidth consumption, thus resolving the sampling rate degradation caused by high sensor density.
2Measurement precision
If all telemetric signals are sampled at high rates to improve fault detection accuracy, then measurement precision is improved, but bandwidth consumption increases beyond system limits
Solution Approach 1:
The patent extracts the redundant information from correlated telemetric signals by identifying that signals within correlated groups contain overlapping information. By sampling only one signal per group and using regression to reconstruct others, the system extracts and eliminates redundant bandwidth consumption while preserving fault detection precision through the regression estimation process.
Solution Approach 2:
The patent creates reconstructed copies of omitted telemetric signals through nonlinear nonparametric regression techniques. Instead of physically transmitting all signals at full rate, the system creates accurate estimates (copies) of the omitted signals from sampled signals, maintaining measurement precision while reducing actual bandwidth consumption to feasible levels.
3Quantity of substance
If sampling rate is reduced to fit bandwidth constraints, then bandwidth consumption is reduced, but fault detection capability deteriorates
Solution Approach 1:
The patent implements feedback through regression analysis where the sampled signals continuously inform the estimation of omitted signals. The regression model uses the sampled signals as input to generate accurate estimates of the full signal set, providing feedback that maintains fault detection capability even though not all signals are directly sampled at the original rate. This feedback mechanism preserves reliability while adhering to bandwidth constraints.
Data Source
AI summary
Some embodiments provide a system that analyzes telemetry data from a monitored system. During operation, the system obtains the telemetry data as a set of telemetric signals from the monitored system and groups the telemetry data into one or more clusters of correlated telemetric signals from the telemetric signals. Next, the system increases a bandwidth associated with monitoring the telemetric signals. To increase the bandwidth, the system omits one or more of the correlated telemetric signals from each of the clusters during sampling of the telemetric signals and estimates the omitted correlated telemetric signals by applying a nonlinear, nonparametric regression technique to the sampled telemetric signals.