Taras Maksymyuk | Lviv Polytechnic National University (original) (raw)

Papers by Taras Maksymyuk

Scientific Reports

In this paper, we propose a solution to resolve the limitation of deep CNN models in real-time ap... more In this paper, we propose a solution to resolve the limitation of deep CNN models in real-time applications. The proposed approach uses multi-threshold binarization over the whole multi-spectral remote sensing image to extract the vector of discriminative features for classification. We compare the classification accuracy and the training time of the proposed approach with ResNet and Ensemble CNN models. The proposed approach shows a significant advantage in accuracy for small datasets, while keeping very close recall score to both deep CNN models for larger datasets. On the other hand, regardless of the dataset size, the proposed multi-threshold binarization provides approximately 5 times lower training and inference time than both ResNet and Ensemble CNN models.

In this paper we analyze the current state of 3G mobile networks in Ukraine and study the prospec... more In this paper we analyze the current state of 3G mobile networks in Ukraine and study the prospects and ways of introducing IoT services based on 4G/5G technologies. We improve architecture of LTE mobile network to provide IoT services through the allocation of narrowband spectrum and the transfer of functions from the eNodeB base station to IoT controller. We propose narrowband spectrum allocation method for IoT in LTE spectrum to reduce inter-cell interference impact on the signal-to-noise ratio at the edges of cells. We also propose method of traffic prioritization to ensure IoT E2E QoS in a heterogeneous LTE/IoT network.

The goal of the metaverse is to provide a synchronized interoperability of real and virtual world... more The goal of the metaverse is to provide a synchronized interoperability of real and virtual worlds across multiple domains and applications. The main challenge is to make the metaverse interface both immersive and responsive to user actions. In this paper, we propose a new concept called a distributed wearable smartphone, which provides a synergistic integration of head-mounted displays, wearable sensors and other smart devices to translate user context and actions across real and virtual metaverse domains. The proposed architecture is flexible and can be composed from any set of devices, depending on the target application. The key benefits of the proposed approach are the realistic 3D representation and the semivirtual user experience, which can be extended to a wide range of different applications in industry, education, healthcare and entertainment.

In 5G communication systems, one key feature is to establish viable spectrum trading platform, wh... more In 5G communication systems, one key feature is to establish viable spectrum trading platform, where would the operators and end-users negotiate about the spectrum availability, quality and finally the price over a short timescale. Our motivation is to investigate the impact of the different spectrum pricing strategies on the operator's strategic decisions in terms of the Sharpe ratio indicator. To be able to numerically quantify Sharpe ratio, we constructed the agent-based model of dynamic spectrum access (DSA) market, where the operators lease the frequency resources from the spectrum exchange server (SpecEx) and end-users are capable of the operator switching based on the preferable utility and price. The numerical experiments suggest that the strategic decision of the operators are influenced by the pricing strategy adopted on the retail market and at the same time the operators adopting different pricing strategies can be classified as either risk-seeking or risk-averse.

2021 International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME), Oct 7, 2021

The COVID-19 pandemic has affected the worldwide health and economy eco-system and remains one of... more The COVID-19 pandemic has affected the worldwide health and economy eco-system and remains one of the key global disturbances even in the presence. A promising solution partially suppressing the fast COVID spread in the dense areas is the digital contact tracing of the infected persons and those who they encounter. Digital contact tracing is an effective solution, whilst few notable drawbacks must be addressed. In general, we are facing both, technological and legal issues, when reconstructing the precise trajectories of the infected persons. The former one stems from the fact that there is commonly a need for personal intervention (specific application, GPS locator, Bluetooth), the latter one deals with the legal aspects of revealing the personal information to the governments/3rd parties. Technological limitations (battery draining, loss of GPS signal, etc.) limits the accuracy of trajectories, especially in dense urban areas. In this paper, we use the autonomous machine learning framework based on the application of the generative adversarial networks (GAN) that is capable of accurate reconstruction of the person trajectories identifying as far as 20% more contacts. Later, this specific person-sensitive information is processed in the private operators' database and blockchain platform is used to securely store healthiness information. The proposed concept facilitates the retrieval of accurate information on the trail of the COVID transparently, being limpid towards privacy and confidentiality concerns.

2018 14th International Conference on Advanced Trends in Radioelecrtronics, Telecommunications and Computer Engineering (TCSET), Feb 1, 2018

Modern optical transport networks are currently facing an unprecedented traffic growth driven by ... more Modern optical transport networks are currently facing an unprecedented traffic growth driven by rapid development of cloud technologies, Internet of Things and ubiquitous computing. The global data volume doubles every two year, requiring urgent improvement of the transport infrastructure around the world. In this paper, we propose a new algorithm for data Hows management in optical label switched networks. Unlike existing solutions, our algorithm provides the intelligence of scheduling and quality control functionality by using machine learning techniques. Intelligence, introduced to the network, improves the accuracy of scheduling and overall performance. Although, initially our algorithm does not provide the near optimal performance like many other approaches, it is able to improve over time by learning from previous experience.

IEEE Communications Magazine, Aug 23, 2020

2022 IEEE 16th International Conference on Advanced Trends in Radioelectronics, Telecommunications and Computer Engineering (TCSET), Feb 22, 2022

International Journal of Distributed Sensor Networks, Mar 1, 2022

The challenge of dynamic traffic demand in mobile networks is tackled by moving cells based on un... more The challenge of dynamic traffic demand in mobile networks is tackled by moving cells based on unmanned aerial vehicles. Considering the tremendous potential of unmanned aerial vehicles in the future, we propose a new heuristic algorithm for coverage optimization. The proposed algorithm is implemented based on a conditional generative adversarial neural network, with a unique multilayer sum-pooling loss function. To assess the performance of the proposed approach, we compare it with the optimal core-set algorithm and quasi-optimal spiral algorithm. Simulation results show that the proposed approach converges to the quasi-optimal solution with a negligible difference from the global optimum while maintaining a quadratic complexity regardless of the number of users.

Journal of Computational Science, Apr 1, 2023

2018 14th International Conference on Advanced Trends in Radioelecrtronics, Telecommunications and Computer Engineering (TCSET), Feb 1, 2018

In this paper, the centralized dynamic resources allocation method for LTE networks is proposed. ... more In this paper, the centralized dynamic resources allocation method for LTE networks is proposed. This method provides dynamic resource allocation according to the users' requirements by small cell virtualization and resource reusing to avoid co-channel interference. The additional novelty of our proposed approach is the additional IoT (Internet of Things) slice overlaying the traditional cellular coverage. This slice provides the transmission of the IoT data by using resource elements, which have not been utilized by mobile users. Thus, important IoT data is transmitted in opportunistic manner, while maintaining the quality of service requirements and throughput demands of mobile users. Resource allocation in proposed method is handled by centralized controller based on C-RAN (cloud radio access network) architecture. Proposed approach improves the bandwidth utilization of the small cells network.

arXiv (Cornell University), Aug 14, 2023

The proliferation of technologies, such as extended reality (XR), has increased the demand for hi... more The proliferation of technologies, such as extended reality (XR), has increased the demand for high-quality threedimensional (3D) graphical representations. Industrial 3D applications encompass computer-aided design (CAD), finite element analysis (FEA), scanning, and robotics. However, current methods employed for industrial 3D representations suffer from high implementation costs and reliance on manual human input for accurate 3D modeling. To address these challenges, neural radiance fields (NeRFs) have emerged as a promising approach for learning 3D scene representations based on provided training 2D images. Despite a growing interest in NeRFs, their potential applications in various industrial subdomains are still unexplored. In this paper, we deliver a comprehensive examination of NeRF industrial applications while also providing direction for future research endeavors. We also present a series of proof-ofconcept experiments that demonstrate the potential of NeRFs in the industrial domain. These experiments include NeRFbased video compression techniques and using NeRFs for 3D motion estimation in the context of collision avoidance. In the video compression experiment, our results show compression savings up to 48% and 74% for resolutions of 1920x1080 and 300x168, respectively. The motion estimation experiment used a 3D animation of a robotic arm to train Dynamic-NeRF (D-NeRF) and achieved an average peak signal-to-noise ratio (PSNR) of disparity map with the value of 23 dB and an structural similarity index measure (SSIM) 0.97.

The end-user mobility patterns play a key role in the process of 5G network design. During tracki... more The end-user mobility patterns play a key role in the process of 5G network design. During tracking end-users via GPS, errors can appear. While we are still constrained with a very limited number of commercially available trajectory datasets, a possible solution is to extend an existing dataset using a generative adversarial network (GAN). Our previous work showed that the utilization of GAN in a form of artificial trajectory generator is possible but not flawless as it introduces the issue with unreasonable gaps between trajectory’s consecutive GPS coordinates. To overcome this issue with a generated dataset and with authentic data as well a special type of GAN called conditional GAN can be used. By leveraging this approach we are not only able to generate a potentially unlimited number of new data samples but as well to correct the existing ones. The number of missing datapoints in the trajectory can go as low as 95% of all points. This artificial intelligence approach has the potential to be used in various use cases where trajectory data are defective and need to be corrected.

2018 International Conference on Information and Telecommunication Technologies and Radio Electronics (UkrMiCo)

Modern 5G networks require complete rethinking of the network management and self-organizing func... more Modern 5G networks require complete rethinking of the network management and self-organizing functionality in order to provide extremely high quality of user experience across multi-tier coverage with different cells radius. In this paper, we propose a novel approach for the small cells coverage planning and performance optimization for multi-tier heterogeneous network based on the artificial intelligence (AI). We develop the AI plane in a way that knowledge is derived there and delivered to the each local SDN controller in a simplified form. Proposed approach allows to effectively train deep neural networks by using generative adversarial networks (GAN) for various topologies in conditions of limited amount of real data about network behavior and performance.

2022 IEEE 16th International Conference on Advanced Trends in Radioelectronics, Telecommunications and Computer Engineering (TCSET)

2022 IEEE 16th International Conference on Advanced Trends in Radioelectronics, Telecommunications and Computer Engineering (TCSET)

2021 IEEE 4th International Conference on Advanced Information and Communication Technologies (AICT), 2021

The end-user mobility patterns play a key role in the process of 5G network design. During tracki... more The end-user mobility patterns play a key role in the process of 5G network design. During tracking end-users via GPS, errors can appear. While we are still constrained with a very limited number of commercially available trajectory datasets, a possible solution is to extend an existing dataset using a generative adversarial network (GAN). Our previous work showed that the utilization of GAN in a form of artificial trajectory generator is possible but not flawless as it introduces the issue with unreasonable gaps between trajectory’s consecutive GPS coordinates. To overcome this issue with a generated dataset and with authentic data as well a special type of GAN called conditional GAN can be used. By leveraging this approach we are not only able to generate a potentially unlimited number of new data samples but as well to correct the existing ones. The number of missing datapoints in the trajectory can go as low as 95% of all points. This artificial intelligence approach has the potential to be used in various use cases where trajectory data are defective and need to be corrected.

International Journal of Distributed Sensor Networks, 2022

The challenge of dynamic traffic demand in mobile networks is tackled by moving cells based on un... more The challenge of dynamic traffic demand in mobile networks is tackled by moving cells based on unmanned aerial vehicles. Considering the tremendous potential of unmanned aerial vehicles in the future, we propose a new heuristic algorithm for coverage optimization. The proposed algorithm is implemented based on a conditional generative adversarial neural network, with a unique multilayer sum-pooling loss function. To assess the performance of the proposed approach, we compare it with the optimal core-set algorithm and quasi-optimal spiral algorithm. Simulation results show that the proposed approach converges to the quasi-optimal solution with a negligible difference from the global optimum while maintaining a quadratic complexity regardless of the number of users.

2018 28th International Conference Radioelektronika (RADIOELEKTRONIKA), 2018

In recent years, there is increasing demand for frequency spectrum due to emerging new multimedia... more In recent years, there is increasing demand for frequency spectrum due to emerging new multimedia applications or concept of the Internet of Things. Nowadays, a strongly discussed topic is the concept of Dynamic Spectrum Access (DSA). It is able to increase efficiency of the licensed parts of the frequency spectrum that are not sufficiently utilized. This paper is devoted to the techno-economical aspects of the DSA in the cognitive radio networks with an emphasis on spectrum trading process. The approach of open access network allows to consider leasing of frequency spectrum by operators as a risky investment opportunity. Our motivation is to investigate impact of interest rate on operators' investment decisions in the spectrum market. The agent-based model was constructed to simulate functioning of the spectrum market. In general, investors are characterized by risk-aversion, which degree changes with respect to wealth. The composition of operators' portfolios is influenced by heterogeneity in sense of different budget constraints. The numerical results of simulations show significant impact of these parameters on behavior of market participants.

Scientific Reports

In this paper, we propose a solution to resolve the limitation of deep CNN models in real-time ap... more In this paper, we propose a solution to resolve the limitation of deep CNN models in real-time applications. The proposed approach uses multi-threshold binarization over the whole multi-spectral remote sensing image to extract the vector of discriminative features for classification. We compare the classification accuracy and the training time of the proposed approach with ResNet and Ensemble CNN models. The proposed approach shows a significant advantage in accuracy for small datasets, while keeping very close recall score to both deep CNN models for larger datasets. On the other hand, regardless of the dataset size, the proposed multi-threshold binarization provides approximately 5 times lower training and inference time than both ResNet and Ensemble CNN models.

In this paper we analyze the current state of 3G mobile networks in Ukraine and study the prospec... more In this paper we analyze the current state of 3G mobile networks in Ukraine and study the prospects and ways of introducing IoT services based on 4G/5G technologies. We improve architecture of LTE mobile network to provide IoT services through the allocation of narrowband spectrum and the transfer of functions from the eNodeB base station to IoT controller. We propose narrowband spectrum allocation method for IoT in LTE spectrum to reduce inter-cell interference impact on the signal-to-noise ratio at the edges of cells. We also propose method of traffic prioritization to ensure IoT E2E QoS in a heterogeneous LTE/IoT network.

The goal of the metaverse is to provide a synchronized interoperability of real and virtual world... more The goal of the metaverse is to provide a synchronized interoperability of real and virtual worlds across multiple domains and applications. The main challenge is to make the metaverse interface both immersive and responsive to user actions. In this paper, we propose a new concept called a distributed wearable smartphone, which provides a synergistic integration of head-mounted displays, wearable sensors and other smart devices to translate user context and actions across real and virtual metaverse domains. The proposed architecture is flexible and can be composed from any set of devices, depending on the target application. The key benefits of the proposed approach are the realistic 3D representation and the semivirtual user experience, which can be extended to a wide range of different applications in industry, education, healthcare and entertainment.

In 5G communication systems, one key feature is to establish viable spectrum trading platform, wh... more In 5G communication systems, one key feature is to establish viable spectrum trading platform, where would the operators and end-users negotiate about the spectrum availability, quality and finally the price over a short timescale. Our motivation is to investigate the impact of the different spectrum pricing strategies on the operator's strategic decisions in terms of the Sharpe ratio indicator. To be able to numerically quantify Sharpe ratio, we constructed the agent-based model of dynamic spectrum access (DSA) market, where the operators lease the frequency resources from the spectrum exchange server (SpecEx) and end-users are capable of the operator switching based on the preferable utility and price. The numerical experiments suggest that the strategic decision of the operators are influenced by the pricing strategy adopted on the retail market and at the same time the operators adopting different pricing strategies can be classified as either risk-seeking or risk-averse.

2021 International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME), Oct 7, 2021

The COVID-19 pandemic has affected the worldwide health and economy eco-system and remains one of... more The COVID-19 pandemic has affected the worldwide health and economy eco-system and remains one of the key global disturbances even in the presence. A promising solution partially suppressing the fast COVID spread in the dense areas is the digital contact tracing of the infected persons and those who they encounter. Digital contact tracing is an effective solution, whilst few notable drawbacks must be addressed. In general, we are facing both, technological and legal issues, when reconstructing the precise trajectories of the infected persons. The former one stems from the fact that there is commonly a need for personal intervention (specific application, GPS locator, Bluetooth), the latter one deals with the legal aspects of revealing the personal information to the governments/3rd parties. Technological limitations (battery draining, loss of GPS signal, etc.) limits the accuracy of trajectories, especially in dense urban areas. In this paper, we use the autonomous machine learning framework based on the application of the generative adversarial networks (GAN) that is capable of accurate reconstruction of the person trajectories identifying as far as 20% more contacts. Later, this specific person-sensitive information is processed in the private operators' database and blockchain platform is used to securely store healthiness information. The proposed concept facilitates the retrieval of accurate information on the trail of the COVID transparently, being limpid towards privacy and confidentiality concerns.

2018 14th International Conference on Advanced Trends in Radioelecrtronics, Telecommunications and Computer Engineering (TCSET), Feb 1, 2018

Modern optical transport networks are currently facing an unprecedented traffic growth driven by ... more Modern optical transport networks are currently facing an unprecedented traffic growth driven by rapid development of cloud technologies, Internet of Things and ubiquitous computing. The global data volume doubles every two year, requiring urgent improvement of the transport infrastructure around the world. In this paper, we propose a new algorithm for data Hows management in optical label switched networks. Unlike existing solutions, our algorithm provides the intelligence of scheduling and quality control functionality by using machine learning techniques. Intelligence, introduced to the network, improves the accuracy of scheduling and overall performance. Although, initially our algorithm does not provide the near optimal performance like many other approaches, it is able to improve over time by learning from previous experience.

IEEE Communications Magazine, Aug 23, 2020

2022 IEEE 16th International Conference on Advanced Trends in Radioelectronics, Telecommunications and Computer Engineering (TCSET), Feb 22, 2022

International Journal of Distributed Sensor Networks, Mar 1, 2022

The challenge of dynamic traffic demand in mobile networks is tackled by moving cells based on un... more The challenge of dynamic traffic demand in mobile networks is tackled by moving cells based on unmanned aerial vehicles. Considering the tremendous potential of unmanned aerial vehicles in the future, we propose a new heuristic algorithm for coverage optimization. The proposed algorithm is implemented based on a conditional generative adversarial neural network, with a unique multilayer sum-pooling loss function. To assess the performance of the proposed approach, we compare it with the optimal core-set algorithm and quasi-optimal spiral algorithm. Simulation results show that the proposed approach converges to the quasi-optimal solution with a negligible difference from the global optimum while maintaining a quadratic complexity regardless of the number of users.

Journal of Computational Science, Apr 1, 2023

2018 14th International Conference on Advanced Trends in Radioelecrtronics, Telecommunications and Computer Engineering (TCSET), Feb 1, 2018

In this paper, the centralized dynamic resources allocation method for LTE networks is proposed. ... more In this paper, the centralized dynamic resources allocation method for LTE networks is proposed. This method provides dynamic resource allocation according to the users' requirements by small cell virtualization and resource reusing to avoid co-channel interference. The additional novelty of our proposed approach is the additional IoT (Internet of Things) slice overlaying the traditional cellular coverage. This slice provides the transmission of the IoT data by using resource elements, which have not been utilized by mobile users. Thus, important IoT data is transmitted in opportunistic manner, while maintaining the quality of service requirements and throughput demands of mobile users. Resource allocation in proposed method is handled by centralized controller based on C-RAN (cloud radio access network) architecture. Proposed approach improves the bandwidth utilization of the small cells network.

arXiv (Cornell University), Aug 14, 2023

The proliferation of technologies, such as extended reality (XR), has increased the demand for hi... more The proliferation of technologies, such as extended reality (XR), has increased the demand for high-quality threedimensional (3D) graphical representations. Industrial 3D applications encompass computer-aided design (CAD), finite element analysis (FEA), scanning, and robotics. However, current methods employed for industrial 3D representations suffer from high implementation costs and reliance on manual human input for accurate 3D modeling. To address these challenges, neural radiance fields (NeRFs) have emerged as a promising approach for learning 3D scene representations based on provided training 2D images. Despite a growing interest in NeRFs, their potential applications in various industrial subdomains are still unexplored. In this paper, we deliver a comprehensive examination of NeRF industrial applications while also providing direction for future research endeavors. We also present a series of proof-ofconcept experiments that demonstrate the potential of NeRFs in the industrial domain. These experiments include NeRFbased video compression techniques and using NeRFs for 3D motion estimation in the context of collision avoidance. In the video compression experiment, our results show compression savings up to 48% and 74% for resolutions of 1920x1080 and 300x168, respectively. The motion estimation experiment used a 3D animation of a robotic arm to train Dynamic-NeRF (D-NeRF) and achieved an average peak signal-to-noise ratio (PSNR) of disparity map with the value of 23 dB and an structural similarity index measure (SSIM) 0.97.

The end-user mobility patterns play a key role in the process of 5G network design. During tracki... more The end-user mobility patterns play a key role in the process of 5G network design. During tracking end-users via GPS, errors can appear. While we are still constrained with a very limited number of commercially available trajectory datasets, a possible solution is to extend an existing dataset using a generative adversarial network (GAN). Our previous work showed that the utilization of GAN in a form of artificial trajectory generator is possible but not flawless as it introduces the issue with unreasonable gaps between trajectory’s consecutive GPS coordinates. To overcome this issue with a generated dataset and with authentic data as well a special type of GAN called conditional GAN can be used. By leveraging this approach we are not only able to generate a potentially unlimited number of new data samples but as well to correct the existing ones. The number of missing datapoints in the trajectory can go as low as 95% of all points. This artificial intelligence approach has the potential to be used in various use cases where trajectory data are defective and need to be corrected.

2018 International Conference on Information and Telecommunication Technologies and Radio Electronics (UkrMiCo)

Modern 5G networks require complete rethinking of the network management and self-organizing func... more Modern 5G networks require complete rethinking of the network management and self-organizing functionality in order to provide extremely high quality of user experience across multi-tier coverage with different cells radius. In this paper, we propose a novel approach for the small cells coverage planning and performance optimization for multi-tier heterogeneous network based on the artificial intelligence (AI). We develop the AI plane in a way that knowledge is derived there and delivered to the each local SDN controller in a simplified form. Proposed approach allows to effectively train deep neural networks by using generative adversarial networks (GAN) for various topologies in conditions of limited amount of real data about network behavior and performance.

2022 IEEE 16th International Conference on Advanced Trends in Radioelectronics, Telecommunications and Computer Engineering (TCSET)

2022 IEEE 16th International Conference on Advanced Trends in Radioelectronics, Telecommunications and Computer Engineering (TCSET)

2021 IEEE 4th International Conference on Advanced Information and Communication Technologies (AICT), 2021

The end-user mobility patterns play a key role in the process of 5G network design. During tracki... more The end-user mobility patterns play a key role in the process of 5G network design. During tracking end-users via GPS, errors can appear. While we are still constrained with a very limited number of commercially available trajectory datasets, a possible solution is to extend an existing dataset using a generative adversarial network (GAN). Our previous work showed that the utilization of GAN in a form of artificial trajectory generator is possible but not flawless as it introduces the issue with unreasonable gaps between trajectory’s consecutive GPS coordinates. To overcome this issue with a generated dataset and with authentic data as well a special type of GAN called conditional GAN can be used. By leveraging this approach we are not only able to generate a potentially unlimited number of new data samples but as well to correct the existing ones. The number of missing datapoints in the trajectory can go as low as 95% of all points. This artificial intelligence approach has the potential to be used in various use cases where trajectory data are defective and need to be corrected.

International Journal of Distributed Sensor Networks, 2022

The challenge of dynamic traffic demand in mobile networks is tackled by moving cells based on un... more The challenge of dynamic traffic demand in mobile networks is tackled by moving cells based on unmanned aerial vehicles. Considering the tremendous potential of unmanned aerial vehicles in the future, we propose a new heuristic algorithm for coverage optimization. The proposed algorithm is implemented based on a conditional generative adversarial neural network, with a unique multilayer sum-pooling loss function. To assess the performance of the proposed approach, we compare it with the optimal core-set algorithm and quasi-optimal spiral algorithm. Simulation results show that the proposed approach converges to the quasi-optimal solution with a negligible difference from the global optimum while maintaining a quadratic complexity regardless of the number of users.

2018 28th International Conference Radioelektronika (RADIOELEKTRONIKA), 2018

In recent years, there is increasing demand for frequency spectrum due to emerging new multimedia... more In recent years, there is increasing demand for frequency spectrum due to emerging new multimedia applications or concept of the Internet of Things. Nowadays, a strongly discussed topic is the concept of Dynamic Spectrum Access (DSA). It is able to increase efficiency of the licensed parts of the frequency spectrum that are not sufficiently utilized. This paper is devoted to the techno-economical aspects of the DSA in the cognitive radio networks with an emphasis on spectrum trading process. The approach of open access network allows to consider leasing of frequency spectrum by operators as a risky investment opportunity. Our motivation is to investigate impact of interest rate on operators' investment decisions in the spectrum market. The agent-based model was constructed to simulate functioning of the spectrum market. In general, investors are characterized by risk-aversion, which degree changes with respect to wealth. The composition of operators' portfolios is influenced by heterogeneity in sense of different budget constraints. The numerical results of simulations show significant impact of these parameters on behavior of market participants.