Configurable Circular Buffer for Real-Time ML Stream Throughput
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multivariate Machine Learning (ML) systems struggle to keep pace with high-rate data ingestion, limiting real-time streaming anomaly detection to offline batch processing and causing failures due to latency buildup.
Innovation Solution
A configurable circular buffer system that automatically adjusts buffer length and single/dual-buffer arrangements to optimize throughput, enabling real-time ML estimation by iteratively testing configurations to balance latency and data ingestion pace.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data is ingested at high rate for real-time streaming, then throughput is improved, but latency builds up causing processing failures
Solution Approach 1:
The patent implements dynamic buffer configuration that automatically adjusts buffer length and switching behavior based on real-time data ingestion rates and processing speeds. The system transitions from static to dynamic buffer management, allowing the buffer to adapt its parameters (length, switching frequency) to balance throughput and latency requirements, resolving the contradiction between high-rate data ingestion and latency control
Solution Approach 2:
The system changes buffer parameters (length, switching frequency, mode) dynamically based on operational conditions. By adjusting these parameters in response to varying data rates and processing capabilities, the system maintains optimal balance between throughput and latency, preventing latency buildup while sustaining high throughput real-time streaming
2Productivity
If buffer length is increased to handle high-rate data streams, then throughput is improved, but memory usage increases
Solution Approach 1:
The patent implements dynamic buffer configuration that automatically adjusts buffer length and switching behavior based on real-time data ingestion rates and processing speeds. The system transitions from static to dynamic buffer management, allowing the buffer to adapt its parameters (length, switching frequency) to balance throughput and latency requirements, resolving the contradiction between high-rate data ingestion and latency control
3Device complexity
If single-buffer arrangement is used to simplify system structure, then device complexity is reduced, but throughput capability decreases
Solution Approach 1:
The patent implements a dual-buffer arrangement that segments the buffer system into multiple independent buffers. This segmentation allows parallel data ingestion and processing operations, enabling the system to handle high-rate data streams effectively. The segmented buffer structure resolves the contradiction by providing the throughput capability of complex systems while maintaining manageable complexity through modular design
Data Source
AI summary
Systems, methods, and other embodiments associated with automatic configuration of a circular buffer for ingesting a stream and generating ML estimates in real-time are described. In one embodiment, an example method includes loading a stream of multivariate time series observations into a circular buffer at a real-time pace of input from a target asset. The circular buffer is configured with a buffer configuration that specifies buffer length and choice of arrangement as a single-buffer or dual-buffer. The method then adjusts the buffer configuration until generation of machine learning estimates of the multivariate time series observations that are in the circular buffer satisfies a threshold test for generation at the real-time pace. And, at the real time pace, the method loads additional multivariate time series observations into the circular buffer that is in the adjusted configuration and generates additional machine learning estimates of the additional multivariate time series observations.