Signal Channel Identification with Joint Peak Likelihoods

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods struggle to accurately identify channels in signal waveforms with multiple peaks that may originate from one or multiple sources, complicating channel identification processes.

Innovation Solution

A method involving capturing a spectrum snapshot, defining regions, performing peak picking, and using periodic peak difference and sub-carrier similarity functions to determine likelihood values, combining these to identify channels based on total likelihood values.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional peak picking methods are used to identify channels in signal waveforms with multiple peaks, then the process is simpler, but the channel identification accuracy deteriorates when peaks may originate from one or multiple sources

Engineering Contradiction:

Improvechannel identification accuracyVSAvoidchannel identification process complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The channel identification process is segmented into multiple independent likelihood calculations: periodic peak difference likelihood, sub-carrier similarity likelihood, and channel density likelihood. Each segment evaluates a specific characteristic of the signal, and their results are combined to achieve accurate channel identification while maintaining modular, manageable complexity

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces multiple dimensions of analysis beyond simple peak detection: temporal periodicity (peak difference analysis), spectral similarity (sub-carrier comparison), and spatial density (channel spacing). By evaluating signals across these multiple dimensions, the system achieves high accuracy in distinguishing channels from multiple sources

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Measurement precision

If multiple likelihood functions are combined to improve channel identification accuracy, then the measurement precision improves, but the computational complexity increases

Engineering Contradiction:

Improvechannel identification accuracyVSAvoidcomputational power

Core Design Contradiction:

Measurement precisionVSPower

Solution Approach 1:

The patent calculates three distinct likelihood values (periodic peak difference, sub-carrier similarity, and channel density) and combines them, using a comprehensive multi-factor approach that achieves high accuracy through partial evaluation of multiple characteristics rather than exhaustive analysis of all possible signal properties

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS20250305812A1Methods for channel identification using joint feature likehoods

Publication Date: 2025.10.02 RAYTHEON APPLIED SIGNAL TECHNOLOGY INC

AI summary

The likelihood that a region of a spectrum of a captured signal waveform defines a channel is determined. A spectrum snapshot is captured of the signal waveform for a spectrum. A plurality of regions is defined within the spectrum. Each region includes a subset of the signal waveform. A peak picking process is performed for the plurality of regions. A first plurality of peaks is identified within a subset of the signal waveform within a region. A periodic peak difference likelihood function is executed to determine a first likelihood value that the plurality of peaks within the subset includes a periodic frequency. A sub-carrier similarity function also is executed to determine a second likelihood value that the plurality of peaks is similar. The values are combined to determine a total likelihood value that is used to determine whether the region and subset of the signal waveform defines a channel.