NEURAL NETWORKS OF THE FINAL RING BASED ON THE REDUCTION SCHEME OF THE POSITION-MODULAR-CODE TRANSFORMATION

The article studies the problem of creating a neural network of modular computing structures for highperformance expressions in the field of information security. The main attention is paid to the reduction technology of position-modular transformation of scalable integers, which serves as the basis for constructing the so-called neural networks of the finite ring (NNFR). To increase the speed of convergence of the reduction scheme used to reduce the number of elements of the generated sequence of residues, an effective tabular method is proposed. The developed approach makes it possible to reduce the number of iterations of the reduction process to a theoretical minimum. This is achieved through flexible adaptive mechanism check botheration deductions to a special range, allowing a tabular decomposition of its elements into pairs of residues in modules of the modular number system. On the basis of a modified reduction method there was synthesized a fast algorithm and a parallel structure of the NNFR with feedback, which ensures the implementation of the reduction scheme in a time order (S(⌈log₂b⌉+1) +2)t_sum, were S – the number of iterations, b – the bit width of the input number, – the duration of the addition operation of two deductions.

1. Chervjakov N. I., Koljada A. A., Ljahov P. A., Babenko M. G., Lavrinenko I. N., Lavrinenko A. V. Moduljarnaja arifmetika i ee prilozhenija v infokommunikacionnyh tehnologijah. Modular Arithmetic and its Applications in Infocommunication Technologies. Moscow, Fizmatlit Publ., 2017, 400 p. (in Russian).

2. Ananda Mohan P. V. Residue Number Systems: Theory and Applications. Basel, Birghauser, Mathematics, 2016, 351 p.

3. Chervjakov N. I., Evdokimov A. A., Galushkin A. I., Lavrinenko I. N., Lavrinenko A. V. Primenenie iskusstvennyh nejronnyh setej i sistemy ostatochnyh klassov v kriptografii. The Use of Artificial Neural Networks and the Residual Class System in Cryptography. Moscow, Fizmatlit Publ., 2012, 280 p. (in Russian).

4. Injutin S. A. Osnovy moduljarnoj algoritmiki. Fundamentals of Modular Algorithms. Khanty-Mansiysk, Poligrafist Publ., 2009, 347 p. (in Russian).

5. Omandi A., Premkumar B. Residue Number Systems: Theory and Implementation. Singapore, Imperial College Press, 2007, 311 p.

6. Ocokov Sh. A. Sposob organizatsii vysokotochnykh vychislenij v modulyarnoj arifmetike [The way to organize high-precision calculations in modular arithmetic]. Pervaja Mezhdunar. konf. «Parallel'naja komp'juternaja algebra i ee prilozhenija v novyh infokommunikacionnyh sistemah» [First International Conf. "Parallel Computer Algebra and Its Applications in New Infocommunication Systems"]. Stavropol, 20–24 Оkt., 2014. Stavropol, Fabula Publ., 2014, pp. 270–277 (in Russian).

7. Gulang G.-B., Mao K.-Z., Siew C. K., Huang D.-S. Fast modular network implementation support vector machines. IEEE Trans. Neural Networks, 2005, vol. 16, no. 6, pp.1651–1663.

8. Tihonov Je. E., Evdokimov A. A. Programmno-apparatnaja realizacija nejronnyh setej : monografija. Software and Hardware Implementation of Neural Networks. Nevinnomyssk, NIJeUP Publ., 2013, 116 p. (in Russian).

9. Sanches D., Melin P., Castillo O. Optimization of modular granular neural networks using arhierarchical genetic algorithm based on the database comlexcity applied to human recognition. Informations Sciences. Tjuana Institute of Technology, 2015, vol. 309, pp. 73–101.

10. Kondrashjov A. V., Gordenko D. V., Pavljuk D. N. Nejronnaya set' dlya preobrazovaniya chisel, predstavlennykh v pozitsionnom kode, v sistemu ostatochnykh klassov [Neural network for converting the numbers represented in the positional code to the residual class system]. Issledovanija v oblasti estestvennyh nauk [Research in the Field of Natural Sciences], 2015, no. 1 (in Russian). Available at: http://science.snauka.ru/2015/01/8925 (accessed 22.01.2018).

11. Babenko M. G., Chernyh A. N., Kuchukov V. A., Derjabin M. A., Kuchukova N. N. Razrabotka novogo nejrosetevogo metoda vychisleniya modul'nogo umnozheniya v sisteme ostatochnykh klassov [Development of a new neural network method for calculating modular multiplication in a system of residual classes]. Nejrokomp'jutery: razrabotka i primenenie [Neurocomputers: Development and Application], 2016, no. 10, рp. 41–48 (in Russian).

12. Koljada A. A. Obobshhennaya integral'no-kharakteristicheskaya baza modulyarnykh sistem schisleniya [Generalized integral-characteristic base of modular number systems]. Informacionnye tehnologii [Information Technology], 2017, vol. 23, no. 9, pp. 641–649 (in Russian).

13. Chernjavskij A. F., Koljada A. A. Preobrazovanie koda modulyarnoj sistemy schisleniya v obobshhennyj pozitsionnyj kod [Conversion of a modular number system code to a generalized positional code]. Doklady Natsional’noi akademii nauk Belarusi [Doklady of the National Academy of Sciences of Belarus], 2017, vol. 61, no. 4, pp. 26–30 (in Russian).

14. Chervjakov N. I., Evdokimov A. A. Nejronnye seti konechnogo kol'tsa dlya realizatsii porogovykh skhem razdeleniya sekreta [Neural networks of the finite ring for the implementation of threshold separation schemes for secretion]. Nejrokomp'jutery: razrabotka, primenenie [Neurocomputers: Development and Application], 2007, no. 2–3, pp. 45–50 (in Russian).

15. Chervjakov N. I., Spel'nikov A. B., Mezenceva A. F. Nejronnaya set' konechnogo kol'tsa pryamogo rasprostraneniya dlya operatsij na ehllipticheskikh krivykh [Neural network of a finite ring of direct propagation for operations on elliptic curves]. Nejrokomp'jutery: razrabotka, primenenie [Neurocomputers: Development and Application], 2008, no. 1–2, pp. 28–34 (in Russian).

16.