Laboratory of Future Networks: HSE Telecommunications Research Institute Develops 5G/6G Research Testbed
The 5G/6G testbed at the HSE Telecommunications Research Institute is becoming a research platform, an educational laboratory, and a foundation for developing new software components for future networks. It makes it possible not only to observe how a mobile network operates, but also to change its operating conditions and measure the results: data-transmission speed, latency, errors, radio-resource utilisation, and other parameters. Based on the testbed, researchers plan to develop MIMO, O-RAN, xApp, and IAB technologies, as well as experiment with artificial intelligence.
Researchers at the HSE Telecommunications Research Institute have developed an experimental 5G/6G testbed that can be used to deploy a functioning test mobile network and study its behaviour under different conditions. The platform already has the key software components of a 5G network up and running, while researchers are collecting telemetry at different layers of the protocol stack and testing MIMO. The next stage will focus on 6G technologies, including O-RAN, xApp, IAB, realistic radio-channel modelling, and neural-network-based solutions. Ultimately, the testbed is intended to connect a scientific idea, a computer model, and an experimental prototype.
Physically, the testbed is a laboratory complex that can be used to deploy a fully functional test mobile network. It comprises computing servers, network software components, equipment for establishing a radio channel, USRP software-defined radio systems, antennas, user devices, and measurement equipment.
The main components of the testbed already in place are a 5G base station, a user device, and the network core. They can be controlled through a dedicated web interface: components can be started and stopped, radio-channel conditions can be changed, network load can be generated, and changes in communication parameters can be monitored. The screen displays the network status, radio-channel characteristics, data-transmission speed, error rate, signal-to-noise ratio, and other metrics.
At the same time, the testbed is not limited to a single approach to experimentation. Its architecture is modular: for some studies, the ability to work with real radio equipment is more important, while for others, the priority is to model specific scenarios quickly and reproducibly.
According to Evgeny Koucheryavy, Director of the HSE Telecommunications Research Institute, a 5G research platform with a number of software components based on 6G technology has already been developed and is now operational. It includes a 5G core, the OAI gNB software-based base station, the OAI UE software-based user device, as well as tools for radio-channel modelling, traffic generation, and monitoring.
The testbed has already successfully deployed OAI gNB, OAI UE, 5G Core, and FlexRIC—a platform for interacting with an intelligent radio-network controller. The researchers have also tested data collection at different layers of the protocol stack—MAC, RLC, and PDCP—under various types of traffic. This makes it possible to obtain detailed telemetry on how the network transmits data and uses radio resources.
The researchers also separately tested MIMO functionality—a technology that enables the simultaneous transmission of multiple data streams. The experiments were conducted both using USRP equipment and in a software simulator.
Modelling and Experimentation
One of the most illustrative experiments carried out on the testbed involves collecting telemetry from a live 5G network. First, the researchers launch a base station, a user device, and the network core, and then generate different types of traffic. A dedicated application collects data on network performance, including transmission speed, delays at different layers of the protocol stack, the amount of radio resources in use, signal quality, transmission errors, and other parameters.
For example, during intensive data transmission, researchers can observe how the load on the radio channel increases, how many resources are allocated to the user, and how transmission speed and quality change as a result.
‘The value of this experiment is that we are observing not an abstract model, but the operation of a real hardware and software system,’ noted Evgeny Koucheryavy.
The data obtained can be used to develop algorithms that automatically make decisions on managing network resources. The testbed has already been used to test data collection at the MAC, RLC, and PDCP layers under different types of traffic. In particular, the system can provide measurements of user-device throughput, RLC-layer latency, and radio-resource utilisation.
This approach is important because a computer model and a physical experiment serve different purposes. A model makes it possible to test an idea quickly, vary a large number of parameters, and conduct numerous experiments. However, every model is based on certain assumptions. In a real system, there are specific interactions between software components, processing delays, hardware limitations, synchronisation errors, and actual radio-channel characteristics that are difficult to fully account for in advance.
‘Modelling and physical experimentation do not compete with each other—they serve different purposes,’ said Evgeny Koucheryavy.
Having its own experimental infrastructure allows researchers to go through the entire research process: first testing a development in a software-based simulation environment, then integrating it into a functioning network and observing how it behaves under real-world conditions.
For example, a new algorithm may perform well in a mathematical model, but after being integrated into a network it may become clear that the data needed to make a decision does not arrive quickly enough, or that the required control action cannot be implemented through the available interfaces. The testbed makes it possible to identify such limitations at the experimental stage.
Speed, Efficiency, and Intelligent Control
One of the main areas of the testbed's development is the radio interface. MIMO makes it possible to use several antennas simultaneously: instead of a single data transmission path, the network has several parallel paths. This increases the speed and efficiency of communications.
The next step is massive MIMO, in which the number of antennas is significantly increased. This makes it possible to direct radio signals more precisely and serve more users simultaneously. Such technologies can be used not only to speed up data transmission but also to analyse the surrounding environment through radio sensing.
Two other important areas are related to network architecture. O-RAN envisages a more open architecture in which network functions can be separated and additional software applications can be connected to them. An xApp is a small, specialised application for managing or analysing a network. For example, it can monitor network load and connection quality and recommend or implement changes to network parameters.
IAB, in turn, makes it possible to use a mobile network not only to communicate with users but also to connect the infrastructure's own elements. Some base stations or nodes can receive connectivity via a wireless link, eliminating the need to lay a separate cable to each one.
For the current stage of the testbed's development, Evgeny Koucheryavy identifies three areas: MIMO, O-RAN with xApp, and IAB. MIMO is directly related to the development of the radio interface and the expansion of network capabilities; O-RAN and xApp make it possible to move from simply monitoring a network to intelligent control; while IAB enables researchers to explore more flexible and distributed architectures for future networks.
The next stage of the testbed's development will also involve integrating Sionna to enable more realistic radio-channel modelling and conducting experiments with a neural-network-based receiver.
AI for the Network and the Network for AI
A separate line of research focuses on artificial intelligence. As the Director of the HSE Telecommunications Research Institute noted, it is important to distinguish between two areas here.
‘AI for the network’ means that AI helps the network itself operate more efficiently: it analyses telemetry, predicts network load, and optimises the use of radio resources.
‘The network for AI’ is the opposite idea. In this case, the network itself becomes an infrastructure that enables the development and delivery of intelligent services to users.
One practical scenario being considered for the testbed involves IMS and AI agents. IMS is a software platform for creating communications services over a mobile network, including voice communications, multimedia services, and modern intelligent services.
The testbed development plan also includes an IMS emulation environment and a demonstration speech-processing service capable of automatically detecting and masking individual words or fragments of an audio stream. In the future, during an ordinary conversation, the system could process speech in real time and provide additional functions, such as call translation, intelligent assistance for an operator, or automated information processing. For the user, it would look like an ordinary phone call, while the network would provide an additional intelligent service in the background.
Scenarios involving IMS, AI agents and, for example, real-time call translation are therefore being considered as one of the areas for the testbed's further development.
For Research and Education
The testbed is being developed not only as a research tool. It is designed to serve several purposes, ranging from scientific experiments to training and the preliminary testing of applied solutions.
For researchers, it is a platform for testing new network-management algorithms and studying MIMO, Open RAN, IAB, and intelligent network control. For students, it provides an opportunity to move from studying 5G/6G architecture using diagrams to conducting experiments themselves: launching a network, connecting a user device, changing channel parameters, generating network load, and analysing the results.
A set of laboratory exercises has already been developed for the testbed and can be used in the university's educational programmes. They could also become part of a specialised continuing professional education programme.
Finally, the testbed can be used in joint projects with companies as a platform for preliminary testing of new solutions. An algorithm can first be developed and tested in a software environment and then evaluated as part of a functioning experimental system.
Thus, the testbed simultaneously serves as research infrastructure, an educational laboratory, and a potential platform for applied development.
From the Laboratory to Industrial Applications
The potential results of such research could be in demand across several industries.
In industry and robotics, mobile robots, sensors, and video-surveillance systems can generate a heavy load on the network simultaneously. In such scenarios, reliable communications, low latency, and the ability to adapt quickly to changing network loads are essential.
In transport and autonomous systems, reliable data exchange is required by autonomous vehicles, drones, and connected infrastructure. MIMO, distributed networks, and intelligent radio-resource management can be applied in these areas.
Another area is the smart city and new intelligent services. A network can connect a large number of devices while simultaneously providing services related to data processing and artificial intelligence, including AI-agent-based services.
Another specific scenario involves areas where laying cable infrastructure is difficult or prohibitively expensive. Here, IAB is of particular interest, as it makes it possible to explore wireless connections between network infrastructure elements.
However, there are several stages between a scientific idea and industrial application. According to Evgeny Koucheryavy, the process begins with formulating a scientific hypothesis or developing an algorithm, which is then tested mathematically and in a software-based simulation environment. The algorithm is subsequently integrated into a functioning network on the testbed, where its impact on speed, latency, errors, connection reliability, and resource utilisation can be measured.
If the experiment produces positive results, a more complete version of the solution is developed—a software module, xApp, network component, or service. The next step is a pilot project, potentially in cooperation with an industrial partner. Only after its effectiveness has been confirmed can the technology be adapted for a specific product, infrastructure, or industry application.
‘The testbed addresses a very important intermediate stage between a scientific idea and implementation. It makes it possible to move from a formula or computer model to an experimentally validated prototype,’ emphasised Evgeny Koucheryavy.
Plans for the Near Future
Over the next year, the researchers plan to achieve a number of measurable results. These include demonstrating that MIMO works as part of the testbed, conducting comparative measurements of throughput under different transmission configurations, measuring average latency and data-transmission reliability, and assessing the proportion of successfully established and maintained communication sessions.
In addition, the researchers plan to develop a set of reproducible MIMO experimental scenarios and a functioning network-telemetry collection system, create an xApp for analysing network-performance parameters, and conduct initial experiments in intelligent network control.
Several other tasks are directly related to the development of 6G technologies. These include deploying and testing a minimum IAB chain, implementing realistic radio-channel modelling that takes a three-dimensional environment into account, and conducting initial tests of a neural-network-based receiver or a functionally similar solution.
Thus, the key outcome of the next stage should be an experimentally validated platform on which the speed, latency, reliability, and efficiency of different algorithms can be measured and compared.
In the future, the testbed will be expanded with 6G technologies and artificial-intelligence tools, both for network management and for creating new intelligent services. At the same time, the 6G standards themselves are still being developed by the international 3GPP standards organisation.
For the HSE Telecommunications Research Institute, this infrastructure serves as a link between several areas at once: fundamental and applied research, specialist training, and prototype development. Its key advantage over a conventional software model is the ability to test a solution not only in simulation but also in a functioning hardware and software system, gradually bringing research results closer to real-world applications.
See also:
Scientists Train Neural Network to Generate Process Plans from 3D Models
Researchers at the HSE FCS AI and Digital Science Institute have developed CAD2TechSpec, a framework that converts 3D models of mechanical parts into machining process plans—step-by-step instructions for machine tools. The solution aims to reduce the time required for the design and preparation of technical process documentation in mechanical engineering, aircraft manufacturing, and other high-tech industries. The study findings have been published in PeerJ Computer Science.
Biologists Discover 'Molecular Fingerprint' of Preeclampsia
Researchers at HSE University employed a new method to model hypoxia in placental cells during pregnancies complicated by preeclampsia and identified molecular markers of tissue hypoxia. Since hypoxia is one of the key mechanisms underlying preeclampsia, these findings are important for a more accurate and timely diagnosis of the disease and for the development of effective treatment methods. The paper has been published in Placenta.
‘Hedgehog’ Versus ‘Relatives’: Researchers Measure How the Brain Responds to Unexpected Words During Natural Speech
Russian neurophysiologists, including researchers from HSE University, have demonstrated the feasibility of using event-related fields (ERFs) to study brain activity during natural speech perception. The researchers showed that this approach can be applied not only to individual words but also to continuous speech. Their findings indicate that words whose meanings differ significantly from the preceding context require longer processing times. The study also reveals that the brain processes function words in two stages: first, it identifies their grammatical role and then uses this information to predict the next word. The study has been published in Frontiers in Human Neuroscience.
Hybrid Intelligence: Competencies in the Age of AI Discussed at Technoprom-2026
Artificial intelligence is not creating new professions, but rather transforming the nature of existing ones. This was the conclusion reached by participants in the panel session ‘Hybrid Intelligence: Digital and Human Drivers of Development,’ organised by the Institute for Statistical Studies and Economics of Knowledge (ISSEK) at HSE University as part of the 13th International Forum of Technological Development (Technoprom-2026). The experts discussed how the nature of work is changing, which skills are becoming increasingly sought after, and what prevents companies from fully capitalising on new technologies.
Scientists Develop New Solution for 6G Communication Systems
A terahertz neuromorphic circuit developed by scientists at HSE University could make 6G communication systems both more accurate and energy-efficient. The circuit enables indoor tracking of mobile devices with an accuracy of up to 99%. The results were presented at PIERS 2026, an international symposium on Photonics and Electromagnetism held in China.
Scientists Develop Algorithm for More Reliable Processors in Data Centres
Researchers from HSE MIEM and Samara University have developed the LRF-3D algorithm to automatically bypass idle nodes in three-dimensional networks-on-chip. Thanks to its hierarchical architecture, the algorithm outperforms existing solutions in both speed and path accuracy, improving processor reliability for use in data centres, supercomputers, and AI computing. The source code and test results are publicly available.
Researchers Rank Recommendation Algorithms Using Sports Tournament Model
Researchers from the AI and Digital Science Institute at the HSE Faculty of Computer Science have developed an approach for selecting recommendation algorithms more effectively. Their approach uses pairwise comparisons of algorithms to create a tournament table, with the overall ranking based on their performance across all datasets in the tournament. This can reduce the number of algorithms that need to be tested when developing new services, saving both time and money. The study was presented at the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2026).
Researchers Develop Method for Direct Generation of Regulatory DNA
Researchers at HSE University have developed a model for generating promoters and enhancers—DNA sequences that regulate gene activity. The model works directly with DNA nucleotides, without first transforming them into a continuous numerical representation. This solution could be useful for applications in synthetic biology and gene therapy. The study results were presented at the ICLR 2026 Workshop ‘Generative AI in Genomics (Gen^2): Barriers and Frontiers.’
Researchers at HSE University and Sber Train Neural Networks to Better Predict User Preferences
The HSE FCS AI and Digital Science Institute and Sber have introduced a new architecture for recommendation systems that combines two classes of models, enabling algorithms to better predict users’ interests and needs. A preprint of the paper has been published on arxiv.org and presented at Urban ML.
Physicists Discover What Happens Inside a Stable Vortex
Large vortices with characteristic spiral arms are often observed in the atmosphere and the ocean. Physicists from HSE University have explained how these structures form and why they retain their shape. The researchers found that velocities at points located along the same vortex arc remain correlated even over long distances. At the same time, this correlation weakens rapidly with increasing distance from the vortex centre. These differences help explain the formation of spiral arms and may improve models of atmospheric and oceanic currents. The findings have been published in Physical Review Fluids.