Continuous Telemetry Collection with Differential Privacy Guarantees

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional differential privacy mechanisms fail to provide strong privacy guarantees when data is collected continuously, especially when users' private values change frequently, limiting their applicability in telemetry applications.

Innovation Solution

A facility is introduced that employs 1-bit response mechanisms and α-point rounding in the local model of differential privacy, combined with memoization and output perturbation, to maintain rigorous privacy guarantees during continuous data collection, ensuring users' privacy is protected even with small but frequent changes in their values.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional differential privacy mechanisms are used for single-round data collection, then strong privacy guarantees are provided, but privacy guarantees quickly degrade when data is collected continuously over time

Engineering Contradiction:

Improveprivacy guarantee strengthVSAvoidduration of continuous data collection

Core Design Contradiction:

ReliabilityVSDuration of action of stationary object

Solution Approach 1:

The patent segments the continuous data collection process into discrete rounds, where each round uses a fresh differential privacy mechanism. This segmentation allows the system to maintain strong privacy guarantees in each individual round while accumulating useful data over multiple rounds, preventing the degradation that occurs with continuous collection using a single mechanism

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary actions by pre-processing and aggregating data in controlled intervals before applying the differential privacy mechanism. This preliminary aggregation reduces the frequency of privacy mechanism applications, maintaining privacy guarantees while still capturing temporal trends in the data

Inventive Principle:

Principle #10Preliminary action

2Measurement precision

If data is collected frequently to capture user behavior changes, then accuracy of telemetry data is improved, but privacy leakage increases

Engineering Contradiction:

Improveaccuracy of user behavior measurementVSAvoidprivacy information leakage

Core Design Contradiction:

Measurement precisionVSLoss of information

Solution Approach 1:

The patent dynamically adjusts the differential privacy parameters (such as epsilon and aggregation intervals) based on the collected data characteristics and privacy requirements. This allows the system to maintain high measurement precision when necessary while reducing privacy leakage by adjusting parameters to match the actual needs of the telemetry application

Inventive Principle:

Principle #35Parameter changes

3Productivity

If telemetry data is collected continuously for long-term storage and analysis, then productivity of data analysis is improved, but privacy protection deteriorates over time

Engineering Contradiction:

Improvedata analysis productivityVSAvoidprivacy protection strength

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent segments stored telemetry data into multiple anonymized batches, each protected by its own differential privacy guarantees. This segmentation enables long-term storage and analysis (improving productivity) while maintaining privacy protection through multiple layers of privacy preservation applied at different time points

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary aggregation layer that processes and anonymizes data before long-term storage. This intermediary layer acts as a mediator between continuous data collection and long-term storage, ensuring privacy protection is maintained even as data accumulates over time for productive analysis

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS10776242B2Collection of sensitive data—such as software usage data or other telemetry data—over repeated collection cycles in satisfaction of privacy guarantees

Publication Date: 2020.09.15 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

A facility for reporting on original values on behalf of each of a plurality of users—each falling within a domain partitioned into a plurality of segments of uniform size—is provided. For each of the plurality of users, the facility (a) randomly selects an upward-rounding window size that is smaller than the segment size; (b) for each source value in a domain, randomly determines a mapping of each segment of the domain to a segment identifier value; (c) determines an original value for the user; (d) adds the upward-rounding window size to the user's original value to obtain a window-augmented original value; (e) identifies a segment containing the window-augmented original value; (f) identifies a segment identifier value mapped-to from the identified segment using the determined mapping; and (g) transmits a reporting communication on the user's behalf reporting a value based on the identified segment identifier value.