Radiotekhnika

Comparative analysis of artificial intelligence based on existing chatbots

Yu.L. Golikov, M.V. Yesina, O.A. Kobylianska — Mon, 25 Dec 2023 00:00:00 +0200

Today, artificial intelligence (AI) is rapidly gaining popularity in various sectors, including the corporate world, business circles, and people's daily lives. The application of artificial intelligence in such fields as medicine, banking and government structures is becoming more frequent. Artificial intelligence facilitates data processing, as it occurs without the intervention of human labor and usually ensures the accuracy of the tasks performed. According to statistics, the number of companies using artificial intelligence in their operations is increasing, and many organizations consider artificial intelligence as an important technology to achieve competitive advantage. This scientific study presents a comprehensive analysis of two leading artificial intelligence systems – ChatGPT-4 from the OpenAI company and Bard from the Google AI company. The work also provides an overview of the development of artificial intelligence in various fields and its impact on everyday human life, especially in such vital areas as medicine, finance, public administration, etc. The paper delves into a detailed comparison of different versions of ChatGPT (GPT-3 and GPT-4) by discussing and analyzing their capabilities, improvements, and limitations. The article also discusses the integration of the Bard system with Google services, its unique functionality and recent updates. The purpose of this study is to compare the capabilities of the artificial intelligence systems ChatGP-4T and Bard, highlight their strengths and weaknesses, as well as their practical application. The paper presents the results of comparative testing to evaluate the performance of each model (system) in various tasks, including solving a logical task, writing an essay, analyzing with subsequent suggestions for improving the web-site, and writing HTML/CSS code for a web-page. The results highlight the fact that, despite the recognized advantages of these models, their functional characteristics may sometimes be limited or not meet expectations when performing specific tasks and the choice of system (model) will be adjusted depending on the needs of users.

Analysis of methods for detecting attacks on facial biometric authentication in mobile devices

O.D. Dolhanenko, O.V. Sievierinov, D.O. Viukhin, V.P. Kotsiuba, A.V. Krepko — Mon, 25 Dec 2023 00:00:00 +0200

The object of the research is facial biometric authentication in the context of mobile devices (financial and government applications, etc.) of the Android and IOS platforms. The subject of the research is methods of detecting attacks on biometric authentication by face in mobile devices. The purpose of the work is to investigate the methods of creating and recognizing Deepfake videos, to evaluate the possibility of local recognition of fakes within applications in mobile devices of the Android and IOS platforms.

As a result, the vulnerability of biometric authentication to Deepfake technology was confirmed, methods for recognizing fakes were analyzed, and the effectiveness of using a neural network model for local recognition of Deepfake in a mobile device was experimentally confirmed. Conclusions are drawn about possible ways to optimize the size of the model with an emphasis on maintaining accuracy.

A method for recovering linear block codes over an arbitrary finite field from sets of distorted code words

A.N. Alekseychuk, O.S. Shevchuk — Mon, 25 Dec 2023 00:00:00 +0200

The article is devoted to one of the practically important problems of information security and cryptanalysis, which consists in recovering an unknown linear block code over an arbitrary field from a set of distorted code words. This is a hard computational problem, and the known problem-solving methods are proposed only for codes over the field of two elements and are based on the algorithms for searching words of small weight in (undistorted) linear block codes.

The main result of the article is a method for solving the problem posed, which differs in essence from the known ones and consists in recovering the desired code by solving the LPN (Learning Parity with Noise) problem, namely, recovering the solutions of systems of linear equations with distorted right-hand sides and a random equally probable matrix of coefficients over specified field. The LPN problem is well known from the Theory of Computational Algorithms and Cryptanalysis. It is equivalent to the problem of random linear block code decoding, and the security of many modern post-quantum cryptosystems are based on its hardness.

The proposed method provides an opportunity to apply a wider class of algorithms for recovering linear block codes in comparison with the previously known methods, in particular, algorithms like BKW and also the low weight words search algorithms in co-sets of linear block codes. Moreover, in contrast to previously known ones, the complexity of the proposed method depends linearly on the length of the required code (and increases with increasing of its dimension according to which algorithm for the LPN problem-solving is applied). Thus, the basic parameter determined the complexity of recovering a linear block code is its dimension (not its length), which, in principle, makes it possible to speed up known algorithms for recovering linear block codes from a set of corrupted code words.

Analysis and comparison of the security of electronic signatures based on new quantum-resistant problems

I.D. Gorbenko, Ye.Yu. Kaptol — Mon, 25 Dec 2023 00:00:00 +0200

Due to the development of quantum computers and quantum methods and algorithms, in order to ensure the security of information after the development of cryptographically relevant quantum computers, NIST conducted the NIST PQC competition. As a result of conducting three rounds of NIST PQC, NIST selected 4 candidates for standardization and four candidates for the fourth round (key encapsulation mechanisms BIKE, Classic McEliece, HQC, and SIKE (which the developers considered unreliable)). Due to the fact that selected algorithms are based on the use of lattices and to add diversity to this list through the use of general-purpose signatures, the process of standardizing additional digital signatures for quantum-resistant cryptography has been initiated.

The following types of signatures are considered for the first round of this standardization process: code-based signatures, isogeny signatures, multivariate signatures, symmetric signatures, MPC-in-the-head, and NIST-defined "other" signatures. These "other" digital signatures are mostly based on new and promising post-quantum (quantum-resistant) problems.

The purpose of the work is to analyze and compare candidates for quantum-resistant digital signatures, based on new and promising quantum-resistant problems, resistant to classical and quantum attacks and side-channel attacks. The paper provides comparison of four digital signatures classified by NIST as "other", namely: ALTEQ, eMLE-Sig 2.0, KAZ-SIGN, Xifrat1-Sign.I. For this purpose, the paper presents the basic principles of each of these digital signatures, their main parameters and available at the time of consideration attack vectors. The paper also provides unconditional criteria necessary for comparison. Digital signatures were compared according to such unconditional criteria as: possible lengths of the public key, possible lengths of the personal (secret) key, length of the result of cryptographic algorithm; and conclusions are made regarding the completeness of given comparison and the possibility of further research is highlighted.

Method of encryption in the MST3 cryptosystem based on Automorphisms group of Suzuki's functional field

Y. Kotukh, G. Khalimov, M. Korobchinskyi — Mon, 25 Dec 2023 00:00:00 +0200

This article presents a new implementation of encryption based on MST, focused on generalized Suzuki 2-groups. The well-known MST cryptosystem, based on Suzuki groups, is constructed using a logarithmic signature at the center of the group, leading to a large array of logarithmic signatures. The proposed encryption is based on multi-parameter noncommutative groups, with a focus on the generalized multi-parameter Suzuki 2-group. This approach responds to the progress in the development of quantum computers, which may pose a threat to the security of many open cryptosystems, especially those based on factorization problems and discrete logarithms, such as RSA or ECC. The use of noncommutative groups to create quantum-resistant cryptosystems has been a known approach for the last two decades.

The unsolvable word problem, proposed by Wagner and Magyarik, is used in the field of permutation groups and is key to the development of cryptosystems. Logarithmic signatures, introduced by Magliveras, represent a unique type of factorization suitable for finite groups. The latest version of such an implementation, known as MST3, is based on the Suzuki group. In 2008, Magliveras introduced the LS transitivity limit for the MST3 cryptosystem, and later Swaba proposed an improved version of the cryptosystem, eMST3. In 2018, T. van Trung suggested applying the MST3 approach using strong aperiodic logarithmic signatures for abelian p-groups. Kong and his colleagues conducted a deep analysis of MST3 and noted that due to the absence of publications on the quantum vulnerability of this algorithm, it can be considered a potential candidate for use in the post-quantum era.

The main distinction of the new system is the use of homomorphic encryption to construct logarithmic signature coverings for all group parameters, which improves the secrecy of the cryptosystem, particularly against brute-force attacks.

New approaches for creating effective combined printed circuit boards on thermally conductive substrates with polyimide dielectrics

Mon, 25 Dec 2023 00:00:00 +0200

New approaches to manufacturing improved combined boards on aluminum bases with various thin polyimide dielectrics, including thermally conductive ones, foiled by copper or aluminum foil have been proposed.

Design and technological solutions and methods for manufacturing combined printed circuit boards on aluminum bases using industrial thin thermally conductive polyimide dielectric films with fluoropolymer coatings with thermal conductivity from 0,12 to 0,46 W/(m·K) have been developed. Design and technological solutions and methods for manufacturing combined printed circuit boards on thermally conductive aluminum bases using adhesive-free copper-polyimide and aluminum-polyimide lacquer foil dielectrics with thermal conductivity of varnish polyimide layers from 0,12 up to 1,0 W/(m·K) and more have been developed.

Main chemical, mechanical, electrical end thermal properties of various types of experimental combined printed circuit boards on aluminum bases with thin polyimide dielectrics for use in electronic modules and printed circuit assemblies (including with using Chip-on-board and Chip-on-flex technologies for assembling) have been studied.

Comparison of the output spectrum for four models of devices with crossed fields

M.E. Alfyorov, M.V. Volovenko, I.O. Moshchenko, O.M. Nikitenko — Mon, 25 Dec 2023 00:00:00 +0200

This paper considers four models of crossed-field devices: cylindrical magnetron, cylindrical magnetron diode, planar magnetron diode and planar magnetron. As is known, in crossed-field devices electrons move by cycloid-like trajectories and induced current arises. The spectral components of the induced current for every of the models are calculated. It has been found that the components of the induced current spectrum contain from one (planar magnetron diode) to eight (cylindrical magnetron diode) components. The calculated spectral components have been compared to the spectral components of the generation spectrum of the real device.

Only one component whose frequency coincides with the cyclotron frequency was found in the induced current spectrum of the planar magnetron diode.

The induced current spectrum in a planar magnetron has three components. These frequencies have a large difference with the frequency of magnetron generation for a microwave oven. Therefore, the models of the planar magnetron diode and the planar magnetron are not very suitable for simulating the operation of crossed-field devices.

The induced current spectrum in a cylindrical magnetron diode contains eight components. The third harmonic is closest to the fundamental frequency of the microwave magnetron. In this case, the difference between the frequencies being compared is near 10 %. Therefore, the model of a cylindrical magnetron diode is not very suitable for modeling the operation of crossed-fields devices.

Two components were found in the induced current spectrum of a cylindrical magnetron. The first harmonic is the closest to the fundamental frequency of the microwave magnetron. In this case, the difference between the frequencies being compared does not exceed 3 %. Such a small discrepancy between the calculated and real frequencies can be explained by the presence of a spatially periodic distribution of the electrostatic field. Thus, the cylindrical magnetron model is well suited for simulating the operation of crossed-fields devices.

Application of artificial intelligence method in adaptive antenna system

M.V. Rozhnovskyi, I.Yu. Rozhnovska — Mon, 25 Dec 2023 00:00:00 +0200

The requirements for adaptive antenna systems in modern and future wireless networks of the fifth (5G) and sixth (6G) generations are analyzed. The block diagram of the adaptive antenna system is presented and the basic principle of its operation is described. It is proposed to improve the block diagram of a modern adaptive antenna system by integrating an artificial intelligence module into it. The principle of interaction of the artificial intelligence module with the adaptive antenna system in the block diagram is shown and described. One of the methods of artificial intelligence (machine learning), the intelligent agent, is described and its mathematical model is presented. The possibility of applying the considered method in the cellular environment of a wireless communication network to improve the operation of an adaptive antenna system is shown. An example of the operation of an artificial intelligence module as part of an adaptive antenna system using an intelligent agent method is given. It is shown that, using the machine learning method, an intelligent agent within a single wireless communication cell can create a certain knowledge system capable of understanding and learning, taking into account the patterns of subscribers’ movement within the cell and predicting the direction of movement of a particular subscriber terminal. The resulting knowledge system is formed in an artificial intelligence module, which is included in the block diagram of a modern adaptive antenna system proposed in this paper, and can potentially be used to more accurately control the directional pattern of an adaptive antenna system. The idea proposed in this paper potentially allows us to develop the concept of a smart antenna, as well as to improve the characteristics of adaptive antenna systems, namely, to increase the energy efficiency of these systems by more accurately realizing the directivity characteristics and intelligent control of the radiation pattern petals using artificial intelligence.

Rectennas of electromagnetic power harvesting systems from the surrounding space

V.O. Alieksieiev, D.V. Gretskih, D.S. Gavva, V.G. Lykhograi, I.A. Khan — Mon, 25 Dec 2023 00:00:00 +0200

Progress in the development of rectenna systems for collecting/harvesting and converting the power of electromagnetic (ЕМ) fields created by radio-electronic means of various classes and purposes into direct current (DC) are considered. The article consists of two parts, each of which highlights important aspects of this topic.

The main parameters of rectennas and the mechanisms of power loss in them are considered in the first part of the article. This allows us to understand the physical and technical limitations that have to be faced in the development of effective ЕМ power harvesting systems.

The second part of the article considers typical schemes of rectennas for use in various applications. Features of the technical implementation of single-band rectennas, options for minimizing their dimensions and the construction of electrically small rectennas based on metasurfaces are given. Schemes of constructions of rectenna arrays and their advantages and disadvantages are also presented. Special attention in the article is paid to multi-band and wide-band rectennas. It is noted that such rectennas can store more energy and produce higher output DC power compared to narrowband rectennas. The design features of low-power rectifiers for rectennas of EM power harvesting systems from the surrounding space are considered.

It was determined that the energy characteristics of rectannas are characterized by numerous factors and there are three approaches to their improvement. The first approach is to improve the parameters of individual rectifier elements, the second is to optimize the parameters of individual rectenna elements and the third is to optimize the entire rectenna as a whole.

One way to improve the quality of a medical manipulator for sterile boxes

O.I. Dovnar, V.I. Cherkis, O.M. Zinchenko, O.O. Kadaner — Mon, 25 Dec 2023 00:00:00 +0200

Today, the advanced robot-surgeons and medical manipulators are managed by operators. Among the best recognized systems are: Da Vinci, Robodoc, SMR Surgical Versius. Further development of medical robotic surgery field requires new solutions in the engineering and software spheres. The main issues and directions are remote surgical operations online as well as automation of processes in sterile boxes. To overcome such limitations, the system must be of minimal size, but it should not lose its accuracy and power. A large number of solutions are aimed at a specific characteristic: Robodoc (accuracy and automation), Da Vinci (flexibility and remote control), SMR Surgical Versius (modularity and miniaturization). If the system has high accuracy, small dimensions and the possibility of automation, it will significantly accelerate the development of robotic surgery.

This task can be solved by removing all electronic components from the segments of the manipulator. Such a solution is already available in the robot surgeon Da Vinci, but a cable mechanism is used there at the base of the segments, which does not allow automating the process because of a low accuracy. The solution that will ensure sufficient accuracy and allow automating the processes is using a bevel gear inside the segments. In this way, it is possible to reduce significantly the size and to increase the accuracy (enlarging the number and type of motors), as well as to make the system as modular as possible, which in its turn will simplify the repair and maintenance.

A manipulator prototype was developed and printed on the NodeMcu hardware platform. The main functionality corresponds to the task. The dimensions of the prototype are 324x324x620 mm, but at the same time, like Da Vinci's robotic surgeon, 7 degrees of freedom remained. Nema17 motors have been used, the accuracy has increased significantly and reaches more than 0.01mm Automation research is still ongoing, but the vibration rate has been significantly reduced because all the motors are placed separately. Technical solutions will allow automating work in sterile boxes, ensuring maximum reliability.

Analysis of the frequency-time structure of acoustic noises of unmanned aerial vehicles in the STM32 CubeIDE environment

V.I. Leonidov, V.V. Semenets, A.B. Grigoriev — Mon, 25 Dec 2023 00:00:00 +0200

The formulation of the task of detecting small unmanned aerial vehicles (drones) is presented, the expediency of building a drone detection system in the stm32 cubeide environment based on the principle of reception and analysis of acoustic signals emitted by drones during their flight mission is substantiated.

The study of temporal fluctuations in the period of acoustic signals of a drone is carried out by the method of model-correlation analysis, as a result of which three-dimensional structures are formed: time – period – correlation coefficient of the acoustic signal with the model in the form of a time-limited sinusoidal function.

The resulting structures are formed as matrices of correlation coefficient values.

The members located along the columns are calculated by time shifting the model function along the signal sample. The members in each column are calculated with a constant period of the model function given from a series of values.

It is shown that the correlation coefficients between the rows of the matrices calculated from drone signals are significantly higher than the same values obtained from background noise measurements. The functions showing the change in time of the correlation coefficients between the rows of the time-period matrix structures for drone signals and background noise do not overlap and show a consistently larger difference in correlation coefficients, which allows us to use the correlation coefficient as a feature that classifies the presence of drone signals.

Method for improving the quality of service of request signals in information systems

I.V. Svyd, D.O. Sukhorukov, O.V. Korotich, T.S. Machonis — Mon, 25 Dec 2023 00:00:00 +0200

The work notes that the request information systems are an important component of information support for air traffic control and airspace control systems. Interrogative information systems include: secondary radar systems, national identification radar systems and short-range navigation radio systems. It is shown that the characteristics of the process of servicing request signals in an aircraft transponder have a significant impact on the statistical properties of the response signal flow, which ultimately, as a consequence, affects the efficiency of the entire radar channel for transmitting response signals with an active response. It is pointed out that one of the important characteristics of the process of servicing request signals in an aircraft transponder is its correlation function, based on the analysis of which one can conclude the choice of the type of response signal flow and the structure of the aircraft transponder. This makes it possible to achieve: increasing the probability of service, increasing the probability of detecting an air object, reducing the root-mean-square error in estimating the azimuth of an air object and increasing the reliability of the transmission of flight information from the observed board of an air object. The article analyzes the possibility of improving the quality of request service by managing the flow of request signals in a network of request information systems. It has been shown that the correct choice of repetition periods of request signals in the area of joint operation of request information systems makes it possible to reduce the likelihood of interaction between individual request signals. Also, as a consequence, it makes it possible to increase the noise immunity of the interrogative information systems under consideration that are part of the specified information system.