Neural Network Terminal Location Estimation for Lower Calculation Cost
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional algorithms for estimating the location of a terminal device require excessive calculation, leading to high costs and reduced accuracy in location estimation.
Innovation Solution
Implementing a target prediction model for location estimation, which reduces estimation cost and improves accuracy by using neural networks to determine the terminal device's location.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional algorithms (Chan's algorithm or Taylor expansion) are used for location estimation, then location estimation can be performed, but excessive calculation is required resulting in excessive cost and reduced accuracy
Solution Approach 1:
The patent transforms the location estimation problem from direct coordinate calculation to a two-stage process: first estimating distance parameters using neural networks, then converting to coordinates. This parameter transformation reduces computational complexity while maintaining accuracy by leveraging the neural network's ability to quickly estimate distance from signal characteristics.
Solution Approach 2:
The patent replaces traditional mathematical algorithms (Chan's algorithm, Taylor expansion) with a neural network-based prediction model. This substitution transitions from deterministic mathematical computation to probabilistic pattern recognition, significantly reducing calculation burden while improving accuracy through learned relationships from training data.
2Productivity
If traditional algorithms are used for location estimation, then location can be calculated, but excessive calculation results in excessive cost
Solution Approach 1:
The patent performs preliminary distance estimation using neural networks before final coordinate calculation. The neural network pre-processes signal characteristics to extract distance information, which then serves as input for coordinate transformation. This preliminary action separates the computationally intensive pattern recognition from the straightforward coordinate math, improving overall efficiency.
Solution Approach 2:
The patent substitutes traditional iterative mathematical algorithms with a trained neural network model that provides direct distance estimates. This replacement eliminates the need for complex iterative calculations, significantly improving productivity while reducing computational cost through the neural network's efficient forward propagation.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
A location estimation method, an information transmission method, a terminal device, and a network device are provided. The location estimation method includes the following. First information sent by a network device is received, where the first information is used for determining a target prediction model. A location of a terminal device or information for determining the location of the terminal device is estimated based on the target prediction model, to obtain an estimation result. Instead of using a traditional algorithm to estimate the location of the terminal device, by adopting the location estimation method, it is possible to not only reduce estimation cost of location estimation, but also facilitate improving accuracy of location estimation.