Indoor Location Tracking via Bayesian Channel Impulse Response Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing indoor positioning systems face challenges in accurately tracking communication devices moving through environments with multipath components, which require intensive processing and large datasets for machine learning techniques, and often rely on deterministic range measurements that are difficult to derive.
Innovation Solution
A computer-implemented method that uses channel impulse responses to determine the location of a communication device by maximizing a combination of a priori and conditional distributions, leveraging both line-of-sight and multipath components without requiring exact deterministic ranges, and employing a Bayesian estimation with particle filtering for efficient real-time tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning techniques are used to incorporate multipath components for localization, then measurement precision is improved, but device complexity increases due to intensive processing requirements
Solution Approach 1:
The patent replaces expensive, complex machine learning models with simpler, computationally efficient probabilistic methods. By using particle filters and Bayesian estimation with pre-computed environmental maps, the system achieves accurate localization without the intensive processing requirements of trained neural networks, effectively using 'cheaper' computational approaches for the same task
Solution Approach 2:
The patent performs preliminary processing by pre-computing and storing environmental maps that contain information about multipath propagation paths, reflective surfaces, and expected signal characteristics. This pre-computation phase creates lookup tables and probabilistic models that can be queried efficiently during real-time tracking, avoiding the need for intensive on-the-fly machine learning inference
2Measurement precision
If machine learning techniques are used to incorporate multipath components for localization, then measurement precision is improved, but loss of energy increases due to large dataset requirements and training phases
Solution Approach 1:
The patent eliminates the need for energy-intensive machine learning training phases by using analytical probabilistic methods. The system achieves comparable or superior performance using closed-form Bayesian solutions and particle filters that require minimal computational resources, thereby significantly reducing energy consumption for both training and inference operations
Solution Approach 2:
The patent extracts and utilizes only the essential features needed for localization from the environmental data, storing compact representations in pre-computed maps. By extracting only the relevant geometric and propagation characteristics rather than using full-scale raw datasets, the system reduces memory requirements and processing energy while maintaining localization accuracy
3Ease of operation
If deterministic range measurements are used for positioning, then ease of operation is improved, but measurement precision deteriorates due to difficulty in deriving exact ranges from channel impulse responses
Solution Approach 1:
Instead of trying to extract deterministic range values from complex channel impulse responses (which is difficult and imprecise), the patent inverts the approach by using the full probabilistic channel response information to directly compute probability distributions of location. The system works with uncertainties and probabilities rather than attempting to force precise deterministic values from ambiguous measurements
Solution Approach 2:
The patent transforms the problem from estimating deterministic range parameters to computing probabilistic location distributions. By changing the mathematical representation from point estimates to probability density functions, the system can fully utilize the information contained in channel impulse responses including multipath components, achieving higher precision without sacrificing operational simplicity
Data Source
AI summary
A computer-implemented method for tracking a communication device moving through an indoor environment; the method includes: i) obtaining channel impulse responses, CIRs, between the communication device and anchor devices along multiple propagation paths; ii) determining an a priori distribution of the location of the communication device within the environment; iii) determining from the channel impulse responses conditional distributions of the range-related measurements conditioned on a range between a potential location of the communication device within the a priori distribution and the respective anchor devices along respective line of sight, LoS, and first order propagation paths; and iv) estimating the location of the communication device as the potential location for which a combination of the a priori distribution and the conditional distributions is maximized.