SNR Estimation via Bessel Function Approximation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating the signal-to-noise ratio (SNR) using the maximum likelihood approach are computationally complex due to the need for numerical solutions, especially in cases of non-linear phase or frequency modulation, leading to high numerical complexity and many iteration steps.
Innovation Solution
The method approximates the quotient of modified Bessel functions by dividing them by a term involving a matching factor and the argument, resulting in a quadratic equation with a self-contained analytical solution, allowing for iterative calculation of signal and noise power using conditional probabilities and Bayes' rule to estimate SNR.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the maximum likelihood approach is used for SNR estimation in non-linearly phase- or frequency-modulated signals, then estimation accuracy is improved, but numerical complexity increases significantly requiring many iteration steps
Solution Approach 1:
The patent transforms the original non-linear SNR estimation problem into a quadratic equation by applying specific mathematical transformations and parameter substitutions. This changes the form of the estimation equation from a complex non-linear form requiring iterative solutions to a quadratic form with closed-form solutions, thereby reducing numerical complexity while maintaining estimation accuracy
Solution Approach 2:
The patent employs simplified estimation equations that can be computed quickly and discarded, replacing the need for complex iterative numerical solutions. The quadratic equation approach provides a computationally inexpensive alternative that achieves the same estimation goal without requiring multiple iteration steps
2Measurement precision
If numerical methods like Newton-Raphson are used to solve the non-linear SNR estimation equation, then a solution can be obtained, but the calculation outlay becomes very high
Solution Approach 1:
The patent extracts the essential SNR estimation problem from the complex numerical solution framework and formulates it as a standalone quadratic equation. This extraction allows the problem to be solved independently using simple closed-form mathematics rather than embedding it in an iterative numerical optimization process that consumes significant computational resources
Solution Approach 2:
The patent replaces the mechanical iterative numerical solution process (Newton-Raphson method) with a direct algebraic solution approach. By substituting the iterative computational mechanism with a closed-form quadratic equation solution, the calculation outlay is dramatically reduced while maintaining the same estimation functionality
3Measurement precision
If iterative calculation with many steps is performed for SNR estimation, then accurate results are achieved, but processing time increases
Solution Approach 1:
The patent performs preliminary mathematical transformations to convert the SNR estimation problem into a quadratic form before actual estimation begins. This preliminary action prepares the equations in advance so that during operation, only simple quadratic formula application is needed, eliminating the need for time-consuming iterative calculations during the actual estimation process
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The method involves determining an estimate value for a squared fading amplitude, an estimate value for noise power of a noise signal and conditional probabilities for data symbols that are modulated corresponding to a transmission signal with a used symbol alphabet. The estimate value for a signal-to-noise ratio is made to correspond to a defined convergence criterion. A quotient is replaced by a proximity term from a modified Bessel function 1st genus and 1st order and another modified Bessel function 1st genus and 0th order, where the quotient is emerged in the iteration. Independent claims are also included for the following: (1) a device for estimating a signal-to-noise ratio of a nonlinear phase or frequency modulated received signal (2) a digital storage medium having a set of instructions for executing a signal-to-noise ratio estimating method (3) a computer program comprising a set of instructions for executing a signal-to-noise ratio estimating method.