Cognitive network management architecture and algorithms
We offer our customers research data and solutions for the intelligent management of telecommunications networks with the goal of optimising current and future systems, such as 5G, in order to reduce maintenance costs, enhance the functionality of the system, and improve the quality of service.
Cognitive and autonomic network management functionalities
VTT has over twenty years of experience in research related to machine learning and artificial intelligence. In our research, we apply this know-how in the automation of telecommunications networks, the importance of which will be emphasised particularly in the future, as the networks grow more complex.
The methods we utilise include reinforcement learning and case-based learning solutions, self-organising maps and Bayesian networks, and fuzzy logic applications for complex decision-making. Other key subjects are various types of decision-making architectures in complex systems and their application in, for instance, software defined networking (SDN) and network function virtualisation (NFV) environments.
Event-based network management architecture and supporting decision-making and learning algorithms for planning and testing mobile network management.
The development and testing platform created by VTT is suitable for developing management algorithms for heterogenic mobile networks and testing them in cases involving both one and several operators. The platform is suited to the development of terminal device and network algorithms. Communication between the components takes place via the Distributed Decision Engine (DDE) developed by VTT. The development and testing platform is integrally connected to VTT's Converging Networks Laboratory (CNL) network testing environment, where the functionality of the management algorithms can be tested in an actual hardware environment in different mobile networks. The hardware testing environment comprises terminal devices, base stations and servers that can be configured to meet the needs of the customer or the project.
VTT offers a development and testing platform and algorithm development for the management of terminal devices and the network.
VTT's development and testing environment has been taken forward in several national and international cooperation projects. It is suited to the design and testing of different types of centralised or distributed mobile network algorithms. The development of management algorithms has recently concentrated on the application of cognitive algorithms to mobility management, application quality assurance, and network load balancing. The aim is to use the algorithms to classify and interpret data monitored on the network and terminal devices, thus detecting any anomalies. The decision-making takes into consideration, for example, performance data from the terminal devices and the network, and user-related contextual data (including location, direction and speed, and the mobile services used). In addition to cognitive algorithms, we also use statistical analyses in the interpretation and classification of data and the detection of anomalies.
The development and testing platform is customisable. The services related to its use are offered and configured to suit each customer and user.
You can watch an example of the use of the platform from the video above: