Electromagnetic Emitter Rank Estimation via Polynomial Curve Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for estimating the number of electromagnetic emitters in military telecommunications are computationally intensive and struggle with accurately identifying co-channel emitters, making them unsuitable for immediate threat recognition.
Innovation Solution
A system and method that utilize a movable detection system to measure angles of arrival (AoAs) of electromagnetic signals, estimate their accuracy, create graphs of AoAs versus range or time, and employ polynomial curve fitting to iteratively test hypothesized ranks, calculating hard scores to determine the rank estimation of emitters, while optimizing results through a greedy search.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If unconstrained programs are used to estimate the number of emitters, then the system can handle complex emitter scenarios including co-channel emitters, but the computational complexity increases making immediate threat recognition difficult
Solution Approach 1:
The patent transforms the emitter estimation problem by changing the parameter space from direct emitter counting to polynomial curve fitting. By representing emitter signals as polynomial functions of range and time, the system converts a complex identification problem into a mathematical fitting problem that can be solved more efficiently while maintaining accuracy for co-channel emitters.
Solution Approach 2:
The patent replaces traditional signal processing mechanical systems with a mathematical model-based approach. Instead of using complex filtering and separation algorithms, the system uses polynomial curve fitting to model emitter signals, substituting mechanical signal processing operations with mathematical transformations that are computationally more efficient.
2Reliability
If traditional emitter estimation methods are used, then the system can identify emitters, but the computational intensity is too high for immediate threat recognition
Solution Approach 1:
The patent applies preliminary action by pre-defining the polynomial model structure that will fit the emitter signals. By establishing the mathematical framework in advance with predetermined polynomial orders and fitting procedures, the system eliminates the need for complex real-time analysis, allowing rapid processing once data is collected.
Solution Approach 2:
The system changes the processing approach by transforming emitter identification into a polynomial fitting problem. This parameter transformation converts the identification task into evaluating polynomial coefficients through curve fitting, which is computationally less intensive than traditional spectral analysis or signal separation methods.
3Measurement precision
If polynomial curve fitting with multiple hypothesized ranks is used, then the rank estimation accuracy improves, but the computational steps increase
Solution Approach 1:
The patent segments the emitter estimation problem by dividing it into discrete rank hypotheses (e.g., 1 emitter, 2 emitters, 3 emitters). Each hypothesis is tested independently through polynomial curve fitting, allowing the system to evaluate multiple scenarios systematically. This segmentation enables accurate rank estimation by comparing fit quality across different emitter count assumptions.
Data Source
AI summary
A system and method for rank estimation of electromagnetic emitters is provided. One aspect of the disclosure provides creating a graph of angles of arrival (AoAs) versus range and using a polynomial curve fit against the graph to determine a rank estimation of electromagnetic emitters. Another aspect of the disclosure provides using a search over parameters of the multiple polynomial curve fits, for each hypothesized rank, to optimize the rank estimation results. This search may be a greedy search to improve speed of convergence. Another aspect of the disclosure provides a metric ‘score’ to select the highest probability rank (number of emitters) based on the agreement between the multiple polynomial curve fits and residual AoA errors.