The space segment market is going to open a new scenario made by flexible satellites based on digital payloads as a consequence of technological development, the emergence of new necessities (such as the exponential increase in demand for connectivity and quality of service), and the corresponding commitment to the full replacement of custom hardware solutions by customizable software alternatives.
These new flexible payloads offer unprecedent capabilities such as the possibility of configurable power and bandwidth allocation as the key capabilities in the future 5G network.
The main objective of ATRIA is the prototyping of a data-driven tool, called Artificial Intelligence – Packet Control Software (AI-PCS), which uses Artificial Intelligence (AI) and Machine Learning (ML) to extract information from data in order to autonomously identify the optimum configuration of the available resources for the on-demand service requests.
AI is a field of computer science related with the production of software capable of some sort of intelligence in a variety of tasks ranging from financial forecasting to image classification and natural language understanding. ML is a subfield of AI containing those algorithms that autonomously learn their ‘rules’ from the data used to train them. AI and ML can therefore move the boundaries of current deterministic processes systems, enabling the detection of hidden patterns invisible to humans, the processing of huge amounts of data, and the making decisions in a very time-efficient manner.
The proposed tool is aimed to be a novelty for the ground segment technologies for two main reasons. First, the AI/ML algorithms will support the payload engineer role in terms of payload configuration; second, the AI/ML algorithms and datasets will provide useful information to explore new capabilities for these complex payloads, allowing cost reduction and improvements of user-experience. ATRIA project and in particular the AI-PCS tool is expected to be valuable for satellite operators, as it will give relevant ad highly data-driven insights about congestion prediction and management, as well as an optimization of gateway diversity, allowing thus an improvement of the user experience for its customers.
I. Faro / L. Taddei / P. Lombardo
Pingback: Data to train AI to help us managing and monitoring a telecommunication network - ATRIA
Comments are closed.