S.G. Chulyuk1

1 JSC «Scientific and Production Association «Radioelectronics», named after V.I. Shimko» (Kazan, Russia)

In the creation of modern and promising radio-electronic complexes, an increasing role is played by the trends of integration and integration of sensors of various functional purposes, providing information about the observed objects.

Recently, developers have been greatly interested in the possibility of sharing information from sensors that provide recognition (classification) of objects to increase the reliability of decisions, including decision-making based on indirect information about objects.

One of such examples is the decision to identify an object on the principle of «friend or foe» (identification) based on the combination of information from various sensors about the class of the object in a certain classification system, which may contain indirect information about whether he is a friend or a foe.

The article attempts to assess the impact of the following factors on the efficiency of combining information: selection of the option of converting the information received from the sensors to a form convenient for combining; the ability to obtain information about the quality of functioning of each sensor and an assessment of the degree of inaccuracy of this information; the possibility of obtaining and evaluating the influence of a priori information on the number and types of objects that may be in the area of the sensors. Three variants of information used for combining are considered: combining primary information from sensor inputs; combining private decisions about the class of the object from the output of the sensors; combining private solutions in the categories «friend or foe» from the output of the sensors.

The issues of obtaining and evaluating the accuracy (reliability) of the initial data, which are necessary for use in the considered algorithms for combining information, are considered.

Requirements for the reliability of the classification and description of objects in the selected system of classes are formulated, at which it is possible to achieve the maximum effect from combining information for solving the identification problem.

Chulyuk S.G. Investigation of the potential for the practical implementation of probabilistic algorithms for combining information in radio-electronic complexes. Radioengineering. 2021. V. 86. № 11. P. 41−46. DOI: https://doi.org/10.18127/j00338486-202111-07 (in Russian)

1. Fomin Ya.A., Tarlovskii G.R. Statisticheskaya teoriya raspoznavaniya obrazov. M.: Radio i svyaz. 1986. (in Russian)
2. Fukunaga K. Vvedenie v statisticheskuyu teoriyu raspoznavaniya obrazov. M.: Mir. 1979. (in Russian)
3. Patent na izobretenie RU 2597870 20.09.2016. Integrirovannoe ustroistvo opoznavaniya. Zhironkin S.B., Makarychev A.V., Sobolev S.A. (in Russian)
4. Chervakov V.O., Zhironkin S.B., Zubarev A.S. Pokazateli tekushchego kachestva statisticheskikh reshenii pri ikh ob'edinenii v kompleksnykh sistemakh obnaruzheniya i raspoznavaniya. Uspekhi sovremennoi radioelektroniki. 2016. № 11. (in Russian)
5. Ivanov S.L., Avramov A.V., Tkachenko S.S. Kompleksirovanie bortovykh radioelektronnykh lokatsionnykh sistem v interesakh kosvennoi identifikatsii tselei. Trudy MAI. 2016. № 87. (in Russian)
6. Chulyuk S.G. Nekotorye aspekty prakticheskogo primeneniya v radioelektronnykh kompleksakh veroyatnostnykh algoritmov ob'edineniya informatsii. Radiotekhnika. 2017. № 8. (in Russian)