Unlock instant, AI-driven research and patent intelligence for your innovation.

A Signal Analysis Method Based on Radiation Source Category Pairs

A signal analysis method and category pair technology, applied in the analysis field of radiation source identification, can solve problems such as the influence of data input order and the importance of signal characteristics.

Active Publication Date: 2015-09-02
THE 28TH RES INST OF CHINA ELECTRONICS TECH GROUP CORP
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Unfortunately, the "alarm defect" problem has not been addressed in existing work on signal feature weighting
In addition to the "alarm defect" problem, the filter feature weighting method has the limitation of discretizing the data in advance, while other feature weighting methods, such as EACH and RELIEF, are limited to heuristic mechanisms and are susceptible to data input order. influences

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A Signal Analysis Method Based on Radiation Source Category Pairs
  • A Signal Analysis Method Based on Radiation Source Category Pairs
  • A Signal Analysis Method Based on Radiation Source Category Pairs

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0073] This example is an example of t-test. Assuming that continuous carrier frequency signals of two types of radiation sources are detected, the unit is megahertz (MHz), category 1 includes continuous carrier frequency signals of five radiation sources, and their values ​​are 1001, 1000, 999, 950 and 1050, respectively. Class 2 includes continuous carrier frequency signals of six radiation sources, and their values ​​are 1500, 1280, 1525, 1290, 1010, and 2000, respectively. On the carrier frequency characteristics, a t-test is performed on the radiation source category pair "1-2", and it is judged whether the carrier frequency characteristics can distinguish the radiation source category pair.

[0074] The t test can be realized by various methods and tools, such as the TTEST function of Microsoft Excel. The t-test yielded a two-tailed p-value of 0.024073. Since this p-value is less than 0.05, it is concluded that the carrier frequency characteristics can distinguish the ...

Embodiment 2

[0076] The present embodiment is an example of Chi-square test. Assume that the discrete signal characteristics of two types of radiation sources—the modulation type—are detected, and its values ​​include (a) "frequency modulation", (b) "amplitude modulation" and (c) "phase modulation". Class 1 includes modulation types of 6 emitters with discrete values ​​of "FM", "FM", "FM", "FM", "AM" and "FM". Class 2 includes 6 modulation types of radiating sources with discrete values ​​of "Am", "Am", "Am", "Am", "Phase" and "Phase". On the modulation type characteristics, a chi-square test is performed on the radiation source category pair "1-2", and it is judged whether the modulation type characteristics can distinguish the radiation source category pair.

[0077] The chi-square test can be implemented with the CHITEST function of Microsoft Excel. The chi-square test using the CHITEST function returns a p-value of 0.012, which is less than the p-value threshold of 0.05. Therefore, ...

Embodiment 3

[0079] It is assumed that the collected emitter data come from four different types of airborne radars: Class 1, Class 2, Class 3 and Class 4. Each radiation source collects four different signal characteristic data: "carrier frequency", "repetition frequency", "pulse width" and "modulation type". Among them, "carrier frequency", "repetition frequency" and "pulse width" are continuous signal characteristics, and "modulation type" is discrete signal characteristics. Assume that according to the t test, the set of radiation source category pairs that can be distinguished by the "carrier frequency" feature is {"1-2", "1-3", "1-4", "2-4"}, and the "repeat frequency" feature The set of radiation source category pairs that can be distinguished is {"1-3", "1-4", "2-4"}, and the set of radiation source category pairs that can be distinguished by the "pulse width" feature is {"2-3", "2-4", "3-4"}; According to the chi-square test, the set of radiation source category pairs that can be...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a signal analysis method based on radiation source category pairs. The method comprises the following steps of: building a cover set of the radiation source category pair of each signal characteristic and calculating a cover coefficient of the radiation source category pair of each signal characteristic; screening redundant signal characteristics according to the cover sets and the cover coefficients of the radiation source category pairs; calculating an optimal cover characteristic of each radiation source category pair; calculating weights of the signal characteristics after being screened according to the cover sets and the cover coefficients of the radiation source category pairs; and outputting the screened signal characteristics, the corresponding characteristic weights as well as the optimal cover characteristics for distinguishing each radiation source category pair. By adopting the method, the category distinguishing ability of the signal characteristics is evaluated in an intuitional and quantized manner according to the cover sets and the cover coefficients of the radiation source category pairs; as the redundant signal characteristics are screened, the imbalance of the category distinguishing ability due to simple superposition of the single signal characteristics in the traditional method is avoided; and meanwhile, the characteristic evaluation results are quantized by the weights.

Description

technical field [0001] The invention relates to the analysis field of radiation source identification, in particular to a signal analysis method based on radiation source category pairs, that is, a signal feature screening, weighting and evaluation method. Background technique [0002] At present, the academic community has proposed a variety of methods to measure the category discrimination ability of signal features, such as information gain, information entropy (entropy), chi-square test (chi-square test) and t test (t test), etc. filter method. However, for multi-category recognition tasks (the number of categories is greater than 2), there is a "alarm defect" in these methods alone-the selected signal features can only distinguish some categories, usually the category with the majority of samples, but cannot distinguish other categories. categories, especially those with a small number of samples, resulting in high recognition accuracy of some categories and low recogn...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/00
Inventor 徐欣朱霞王红杰
Owner THE 28TH RES INST OF CHINA ELECTRONICS TECH GROUP CORP