Research areas
Management, orchestration, and control of next-generation networks and services
The department focuses primarily on developments in service and network architectures. The objective is to develop automated and potentially autonomous approaches to orchestrating, controlling, and managing future networks. Department members drive these developments based on the following principles:
- the separation of hardware and software (including virtualization and the sharing of resources and infrastructure, as well as infrastructure, service, and function virtualization in the context of cloud computing and networks),
- loose coupling and separation between services, based on microservices and component-based approaches,
- separation of network management, control, and forwarding planes, in connection with SDN and NFV, enabling dynamic network creation and programmability, and leading to network slicing solutions (Slices as a Service).
The department designs, models, and assesses future service and network architectures based on these principles, as well as network and system modeling, optimization, machine learning, and artificial intelligence.
Advanced 5G/6G networks and edge computing:
RS2M develops intelligent architectures for next-generation networks, including network slicing, automated orchestration, and serverless edge computing. Research focuses on dynamic resource allocation, adaptive pricing, and optimizing ultra-reliable, low-latency communications (URLLC) for critical applications, such as autonomous vehicles.
Distributed artificial intelligence (AI for networks, networks for AI, AI-as-a-service):
The department also works on integrating AI into networks through AI-native networking. Additionally, the department explores the role of networks in delivering AI solutions to different industries and sectors (AI as a service). The automation solutions being developed are based on deep reinforcement learning (DRL), federated learning, and agentic solutions. These solutions have applications in drone trajectory planning, mobility prediction in vehicular networks, and integrating large language models (LLMs) for data analysis and misinformation detection.
Mathematical optimization and game theory:
The department has extensive expertise in solving complex optimization problems by combining graph theory, combinatorics, and linear programming. These methods are applied to the modeling of distributed systems, coalition games for joint infrastructure investment, and the optimization of industrial processes.
Formal methods, testing, and validation
The department's work in formal methods, testing, and validation focuses on testing strategies based on formal models of stateful and stateless systems. These systems include robots, finite-state machines, logic circuits, and formulas expressed in various logics. Specifications are validated against functional and non-functional properties. Then, tests are generated to assess system compliance with specifications, seeking a balance (or trade-off) between complexity and fault coverage. Research also addresses the verification of security properties in services and protocols, as well as the interoperability of network components.
Examples of application areas and sectors
Wireless sensing and healthcare applications:
RS2M is a pioneer in the use of Wi-Fi, mmWave, and ultra-wideband signals for non-invasive vital sign monitoring. The systems developed enable multi-user cardiac and respiratory monitoring, sleep monitoring, and monitoring in medical emergencies.
Blockchain and trustworthy systems:
The department develops Byzantine-fault-tolerant consensus protocols and distributed trust architectures for the industrial IoT, incorporating advanced privacy-preserving and access-control mechanisms.
