Embedded Files
While 5G networks are starting to make their way in our society, the research on 6G networks has already started. A new generation always represents a breakthrough, either in terms of new technology, new architecture and new services. Artificial Intelligence (AI) will probably be a key aspect of 6G networks. In our vision, as expressed in [1], besides being AI-native, 6G networks will move beyond the conventional Shannon legacy to incorporate semantic and goal-oriented aspects. The main idea of semantic communication is that communication takes place to carry meaning and the recovery of meaning is what drives the design of semantic encoders/decoders. Combining knowledge representation and reasoning tools with machine learning algorithms paves the way to build semantic learning strategies enabling current machine learning algorithms to achieve better interpretation capabilities and better contrast adversarial attacks. On the other side, designing goal-oriented communications requires first to formally define the purpose of communication and, as a consequence, to identify the relevant information, i.e. the information strictly necessary to recover the meaning intended by the transmitter or to accomplish a goal. This new vision, once properly formalized, has the potential of achieving a much better trade-off in terms of energy, delay, and learning capabilities. 6G semantic networks can bring semantic learning mechanisms at the edge of the network and, at the same time, semantic learning can help 6G networks to improve their efficiency and sustainability.
Sergio Barbarossa, Danilo Comminiello, Eleonora Grassucci, Francesco Pezone, Stefania Sardellitti, Paolo Di Lorenzo, Semantic Communications Based on Adaptive Generative Models and Information Bottleneck, to appear in the IEEE Communications Magazine, Nov. 2023.
Page updated
Report abuse