POEM Privacy-preserving and Channel-adaptive Smart Task-oriented Semantic Communication
- Coordinator
- The University of Exeter, United Kingdom
- Runs
- 4 August 2025 to 3 August 2027
- Status
- Running
- EU funding
- €276k
- Total cost
- —
- Partners
- 1 organisations in 1 countries
- Type
- Marie Skłodowska-Curie Action
- Grant agreement
- 101212103
What the project is doing
In the project's own words, from its summary on CORDIS.
Integrating Artificial Intelligence (AI) into the sixth-generation (6G) wireless networks generates a promising paradigm, namely task-oriented semantic communications (ToSC), which focuses on task efficiency by transmitting task-relevant information only and thus reducing communication overhead. ToSC is crucial for emerging applications, such as augmented reality, autonomous driving, and smart healthcare, which require frequent, large-scale data exchanges.
However, ToSC faces challenges related to explainability of decision-making, data privacy, and adaptability in dynamic wireless environments. To address these challenges, this project aims to create an innovative, privacy-preserving, and channel-adaptive smart ToSC system. First, groundbreaking transparent neural networks (NNs) will be developed by using model-based learning and deep unfolding techniques, providing interpretable decision-making while maintaining performance.
Second, a novel privacy-preserving framework will integrate adversarial models and encrypted parameter sharing in ToSC, ensuring robust data protection. Third, a pioneering channel-adaptive system will be developed using diffusion-based models and meta-learning techniques, enabling real-time adaptability to varying wireless conditions without retraining the NNs. This new system will be validated through a simulation platform and hardware testbed to ensure optimal performance in real-world scenarios.
The success of this project will deliver groundbreaking advancements in wireless communication and AI, providing explainable, adaptive, and secure ToSC for future 6G networks. By reducing the volume of data transmission, the outcomes of this project will cut down energy consumption and lower carbon emissions. In addition to addressing the key concerns like data privacy and sustainable communication, this project will foster talent development through interdisciplinary training and position the EU as a leader in AI-driven communication technologies.
Who is involved
| Organisation | City | Country | Role |
|---|---|---|---|
| The University of Exeter | Exeter | United Kingdom | Coordinator |
Source: CORDIS, European Commission. Data of 5 Nov 2026, reused under the Commission's reuse policy. View on CORDIS · Grant DOI