Flow Rate Prediction Device for Mixed State Throughput

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing communication throughput prediction devices face reduced accuracy when applied to environments where communication throughput is in a mixed steady and non-steady state, and current methods for estimating mixing ratios are not optimal.

Innovation Solution

A flow rate prediction device and method that discerns between steady and non-steady states using time series data, estimates a mixing ratio using a probabilistic fluctuation model, and constructs a mixed model by combining steady and non-steady models to improve prediction accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a communication throughput prediction device categorizes the state of communication throughput into either a steady state or a non-steady state, then the prediction model can be simplified, but the prediction accuracy is reduced when applied to environments where the communication throughput is in a mixed steady and non-steady state

Engineering Contradiction:

Improveprediction model complexityVSAvoidprediction accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent divides the communication throughput state into two distinct segments: steady state and non-steady state. By creating separate prediction models for each segment and determining which segment the current data belongs to, the system achieves accurate predictions for mixed states without excessive complexity. The segmentation is implemented through stationarity testing that categorizes each time series segment into one of the two states.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent implements a dynamic approach by continuously testing the stationarity of communication throughput data and adapting the prediction model based on the current state. The system transitions between steady and non-steady models dynamically, allowing it to respond to changing network conditions. This dynamic state recognition enables accurate prediction in mixed environments where the network alternates between stable and unstable periods.

Inventive Principle:

Principle #15Dynamics

2Ease of operation

If a smoothing filter is used to calculate the mixing ratio based on stationarity analysis results, then the mixing ratio can be obtained, but the estimation accuracy of the mixing ratio is not optimal

Engineering Contradiction:

Improvemixing ratio calculation simplicityVSAvoidmixing ratio estimation accuracy

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The patent replaces the mechanical smoothing filter approach with a probabilistic model-based estimation method. Instead of mechanically averaging the stationarity results, the system uses probability theory to estimate the mixing ratio by analyzing the distribution of steady and non-steady states in the time series data. This substitution of the estimation mechanism significantly improves accuracy while maintaining computational feasibility.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS10445444B2Flow rate prediction device, mixing ratio estimation device, method, and computer-readable recording medium

Publication Date: 2019.10.15 NEC CORP

AI summary

Communication throughput or fluctuations therein is more accurately predicted. A flow rate prediction device of the present invention is provided with: a stationarity discerning means for discerning whether the state of fluctuation in a flow rate is a steady state or non-steady state, on the basis of flow rate time series data; a mixing ratio estimating means which uses the discernment result, a mixing ratio fluctuation model, and a discernment result observation model, and which estimates the mixing ratio in the flow rate in a designated duration; and a model mixing means for mixing a steady model and a non-steady model, on the basis of the estimated mixing ratio, to construct a mixed model.