Entropy-Based Fingerprinting for Indoor WLAN Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing indoor localization techniques face challenges in non-line-of-sight environments and multipath propagation, with RSS-based fingerprinting suffering from low precision due to signal fluctuations and CIR/CTF-based methods being computationally and storage-intensive.
Innovation Solution
A method and system using entropy estimation of wireless communication channel functions, particularly the channel transfer function, to create a robust fingerprinting structure that reduces computational complexity and storage requirements, enabling accurate localization without the need for matrix manipulation and storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If RSS-based fingerprinting is used for indoor localization, then the implementation is simple and readily available, but the localization precision is low due to signal fluctuations and multipath effects
Solution Approach 1:
The patent transforms the channel impulse response (CIR) into its frequency-domain representation (CTF) and then extracts entropy values as a new feature parameter. This parameter transformation captures multipath information in a compressed form that is both unique for localization and computationally efficient, resolving the contradiction between simplicity and precision.
2Measurement precision
If CIR-based fingerprinting is used to capture multipath information, then localization accuracy is improved, but computational complexity and storage requirements increase significantly
Solution Approach 1:
The patent extracts only the essential information from the complex CIR data by transforming it to the frequency domain and computing entropy values. This extraction process retains the unique multipath characteristics needed for accurate localization while discarding redundant time-domain information that causes computational burden.
Solution Approach 2:
By changing the representation from time-domain CIR to frequency-domain CTF and then to entropy values, the patent reduces the dimensionality and complexity of the data while preserving the essential localization information. This parameter transformation makes the system more efficient without sacrificing accuracy.
3Measurement precision
If CTF-based fingerprinting is used to represent multipath channel information, then localization precision is improved, but storage requirements and computational burden increase
Solution Approach 1:
The patent extracts the essential multipath information from the CTF by computing entropy values, which compresses the data into a more compact representation. This extraction reduces the amount of data that needs to be stored and processed while maintaining the unique characteristics needed for precise localization.
Solution Approach 2:
The transformation from CTF to entropy values changes the data representation into a more compact form. This parameter change reduces the quantity of data required for storage and computation while preserving the essential information for accurate location identification.
4Measurement precision
If pattern recognition algorithms are used to compare fingerprints, then localization accuracy is improved, but processing time and computational cost increase
Solution Approach 1:
By changing the fingerprint representation to entropy values, the patent creates a more compact and distinctive feature set that enables faster pattern recognition. The entropy-based fingerprints require fewer computational operations to compare and match, reducing processing time while maintaining accuracy.
Data Source
AI summary
This invention relates to methods and devices for entropy-based location fingerprinting, in particular for use over wireless local-area networks (WLANs). The invention has particular application in localization for indoor environments. In embodiments of the invention, an entropy-based fingerprint is determined at a number of predetermined locations within the desired area of localization during an off-line phase and subsequently used in an on-line mode to determine the location of a receiver. In particular embodiments, the fingerprint is a vector of entropy estimates of the channel transfer function (CTF) between a mobile terminal and all access points within coverage. The invention seeks to provide a fingerprinting localization solution that has a simplicity of structure, leading to advantages in storage and pattern recognition requirements, and robustness by proving a unique measure of information that is related to the channel experienced at the location of the mobile terminal.