HARDWARE NEURAL NETWORK BASED MEMRISTIVE TITANIUM OXIDE STRUCTURES
Abstract
The paper presents the results of manufacturing, training and research of a hardware neural network prototype implemented as a crossbar array of artificial synapses based on memristor nanostructures of electrochemical titanium oxide. A prototype of a fully connected neural network was developed, consisting of four input electrodes, a crossbar array of 16 artificial synapses based on electrochemical titanium oxide nanostructures and four output electrodes. It is shown that the process of current flow through such a structure fully corresponds to the mathematical model of the neural network. Various implementations of artificial synapses that allow the implementation of negative "weights" of the neural network were analyzed and one of the optimal options was selected. Based on the developed structure, a prototype of a fully connected neural network was manufactured using magnetron sputtering, optical lithography and nanolithography technologies using scanning probe microscopy methods. To train the neural network, an algorithm for switching individual memristors was developed, eliminating parasitic switching of neighboring structures due to the occurrence of leakage current. To demonstrate the operation of the manufactured neural network model, a task of classifying two input signals was proposed. To implement negative "weights", each of the incoming signals was duplicated with negative polarity. It is assumed that the outputs of the trained neural network should register: 1) the excess of the first signal; 2) the excess of the second signal; 3) both high signals. The training and research of the neural network was carried out using the hardware and software complex "Neuro InT", developed by the staff of the Research Laboratory "Neuroelectronics and Memristive Nanomaterials", SFedU. Research of the neural network model showed that all outputs successfully classify incoming signals, maximizing the current through the corresponding outputs for the given input values. The proposed structure can be improved by adding two additional inputs with a constant high positive and negative potential to implement a "shift" during the operation of the neural network. The obtained results can be used in the development of technological foundations for the formation of hardware neural networks based on memristor titanium oxide nanostructures
References
1. Soori M., Arezoo B., Dastres R. Artificial neural networks in supply chain management, a review, Jour-nal of Economy and Technology, 2023, Vol. 1, pp. 179-196.
2. Jaekwang Cha, Shiho Kim CNN Hardware Accelerator Architecture Design for Energy-Efficient AI, Artificial Intelligence and Hardware Accelerators, 2023, pp. 319-357.
3. Abhijit Pandya, Ankur Agarwal, P. K. Kim Low Power Design of the Neuroprocessor, Knowledge-Based Intelligent Information and Engineering Systems, 2003, Vol. 2774, pp. 856-862.
4. Fei Zhang, Mehdi Aghagolzadeh, Karim Oweiss A Fully Implantable, Programmable and Multimodal Neuroprocessor for Wireless, Cortically Controlled Brain-Machine Interface Applications, J Sign Pro-cess Syst,, 2012, Vol. 69, pp. 351-361.
5. Xiaoyang Liu, Zhigang Zeng, Rusheng Ju Design of Memristor-Based Binarized Multi-layer Neural Network with High Robustness, Neural Information Processing. Communications in Computer and In-formation Science, 2024, Vol. 1962, pp. 249-259.
6. Mousam Charan Sahu, Anjan Kumar Jena, Sameer Kumar Mallik, Suman Roy, Sandhyarani Sahoo, et al. Reconfigurable Low-Power TiO2 Memristor for Integration of Artificial Synapse and Nociceptor, ACS Applied Materials & Interfaces, 2023, Vol. 15 (21), pp. 25713-25725.
7. Tominov R., Avilov V., Vakulov Z., Khakhulin D., Ageev O., Valov I., Smirnov V. Forming-Free Resis-tive Switching of Electrochemical Titanium Oxide Localized Nanostructures: Anodization, Chemical Composition, Nanoscale Size Effects, and Memristive Storage, Adv. Electron. Mater., 2022, 2200215.
8. Avilov V.I., Smirnov V.A., Tominov R.V., Sharapov N.A., Avakyan A.A. Atomic force microscopy of titanium oxide nanosize structures resistive switching, Abstract Book of International Conference “Scanning Probe Microscopy”, 2019, pp. 131-132.
9. Avilov V.I., Smirnov V.A., Tominov R.V., Sharapov N.A., Polupanov N.A., Ageev O.A. Phase composi-tion investigation of titanium oxide nanostructures obtained by the local anodic oxidation, IOP Conf. Se-ries: Materials Science and Engineering, 2019, Vol. 699, 012003.
10. Smirnov V.A., Avilov V.I., Tominov R.V., Fedotov A.A., Ageev O.A., Valov I. Memristornye struktury na osnove elektrokhimicheskogo oksida titana dlya RERAM i neyromorfnykh primeneniy [Memristor structures based on electrochemical titanium oxide for RERAM and neuromorphic applications], Nanoindustriya [Nanoindustry], 2021, Vol. 14, pp. 664-665.
11. Smirnov V.A., Tominov R.V., Avilov V.I., Polyakova V.V., Ageev O.A. Issledovanie effekta rezistivnogo pereklyucheniya v ne trebuyushchikh formovki oksidnykh nanorazmernykh strukturakh titana [Study of the effect of resistive switching in oxide nanoscale titanium structures that do not require forming], Izvestiya YuFU. Tekhnicheskie nauki [Izvestiya SFedU. Engineering Sciences], 2019, No. 2 (204), pp. 201-213.
12. Avilov V.I., Smirnov V.A., Sharapov N.A. Razmernyy effekt v memristornykh nanostrukturakh na os-nove oksida titana dlya sozdaniya elementov sistem iskusstvennogo intellekta i sinaptroniki [Size effect in memristor nanostructures based on titanium oxide for creating elements of artificial intelligence and synaptronics systems], Izvestiya YuFU. Tekhnicheskie nauki [Izvestiya SFedU. Engineering Sciences], 2018, No. 2 (196), pp. 34-46.
13. Avilov V.I. Zakonomernosti formirovaniya i proyavleniya rezistivnogo pereklyucheniya v nanostruktu-rakh oksida titana dlya apparatnykh neyronnykh setey [Patterns of formation and manifestation of resis-tive switching in titanium oxide nanostructures for hardware neural networks], Perspektivnye sistemy i zadachi upravleniya: Mater. XIX Vserossiyskoy nauchno-prakticheskoy konferentsii [Advanced control systems and problems: Proceedings of the XIX All-Russian scientific and practical conference], 2024, pp. 419-423.
14. Zhavoronkov L.G., Avilov V.I., Polupanov N.V., Khakhulin D.A., and Smirnov V.A. Fabrication and investigation of a memristive crossbar array artificial synapses based on electrochemical titanium oxide for neuroelectronics, Ferroelectrics, 2024, Vol. 618 (5), pp. 1323-1329.
15. Khakhulin D.A., Dzyuba D.A., Avilov V.I., Smirnov V.A. Sinapticheskie svoystva memristora na osnove TiOx [Synaptic properties of a TiOx-based memristor], Kurchatovskaya mezhdistsiplinarnaya mo-lodezhnaya nauchnaya shkola: Sb. annotatsiy [Kurchatov Interdisciplinary Youth Scientific School: Collection of Abstracts], 2023, 87.
16. Avilov V.I., Varganov V.I., Fedotov A.A., Smirnov V.A. Neyromorfnye struktury v sistemakh RTK [Neu-romorphic structures in RTC systems], Perspektivnye sistemy i zadachi upravleniya: Mater. XVIII Vse-rossiyskoy nauchno-prakticheskoy konferentsii i XIV molodezhnoy shkoly-seminara [Advanced systems and control problems: Proceedings of the XVIII All-Russian Scientific and Practical Conference and XIV Youth School-Seminar], 2023, pp. 173-176.
17. Avilov V.I., Zhavoronkov L.G., Polupanov N.V., Khakhulin D.A., Smirnov V.A. Sinapticheskie ustroystva dlya neyromorfnykh sistem robototekhnicheskikh kompleksov [Synaptic devices for neuro-morphic systems of robotic complexes], Perspektivnye sistemy i zadachi upravleniya: Mater. XVIII Vse-rossiyskoy nauchno-prakticheskoy konferentsii i XIV molodezhnoy shkoly-seminara [Prospective sys-tems and control problems: Proceedings of the XVIII All-Russian scientific and practical conference and XIV youth school-seminar], 2023, pp. 169-173.
18. Tominov R.V., Avilov V.I., Chernenko N.E., Smirnov V.A. Issledovanie memristornogo effekta tonkoy plenki oksida titana dlya iskusstvennykh neyropodobnykh sistem [Study of the memristor effect of a thin titanium oxide film for artificial neuron-like systems], Sb. trudov XIII Vserossiyskoy konferentsii mo-lodykh uchenykh «Nanoelektronika, nanofotonika i nelineynaya fizika [Collection of works of the XIII All-Russian conference of young scientists "Nanoelectronics, nanophotonics and nonlinear physics], 2018, pp. 318-319.
19. Karen’kih O.G., Avilov V.I., Smirnov V.A., Fedotov A.A., Sharapov N.A. and Polupanov N.A. Modelling of local anodic oxidation of titanium oxide nanostructures formation process. IOP Conf. Series: Materi-als Science and Engineering, 2018, Vol. 443, 012013.
20. Avilov Vadim I., Tominov Roman V., Vakulov Zakhar E., Zhavoronkov Lev G., and Smirnov Vladimir A. Titanium oxide artificial synaptic device: Nanostructure modeling and synthesis, memristive cross-bar fabrication, and resistive switching investigation, Nano Research, 2023, Vol. 16, pp. 10222-10233.
21. Avilov Vadim I., Tominov Roman V., Vakulov Zakhar E., Rodriguez Daniel J., Polupanov Nikita V., Smirnov Vladimir A. Nanoscale Titanium Oxide Memristive Structures for Neuromorphic Applications: Atomic Force Anodization Techniques, Modeling, Chemical Composition, and Resistive Switching Properties, Nanomaterials, 2025, Vol. 15 (1), 75.
22. Karen’kih O.G., Avilov V.I., Smirnov V.A., Sharapov N.A., Polupanov N.A. Modeling of titanium oxide nanostructures formation process by local anodic oxidation, Abstract Book of International Conference “Scanning Probe Microscopy”, 2018, 97.
