SemComIIoT

SemComIIoT aims at developing research in the context of semantic communications for Industrial IoT environments 6G large-scale sensing environments. Where 6G applications (such as Metaverse applications) are expected to be based on a massive number of mobile devices, intelligent and with reasoning capabilities. The underlying infrastructure (wireless, cellular) will have to cope with massive volumes of data in real time and in a reliable way. For the specific case of IioT, this implies transmitting data with minimum and maximum latency (below 1ms), achieving low jitter, and zero packet loss, while at the same time handling the so-called mobility gap, across heterogeneous networks, relying on different frequencies.

In contrast to current host-oriented communication networks, semantic communications (SemCom) rely on coding approaches (e.g., network coding) to be able to inject data meaning, instead of raw data toward receivers. SemCom therefore relies on Edge Artificial Intelligence (AI) pre-processing, coupled with coding at the lower network layers (OSI MAC and PHY), to extract relevant meaning to be transmitted, filtering unessential information.

In this context, the partners aim to explore SemCom in the context of Industrial IoT environments. For this purpose, This project mainly includes three research topics: Research on Semantic Information Theory, Design the Semantic Communication System for Industrial IoT Scenarios, Semantic-aware Transmission.

The project shall give rise to the following contributions:

• Objective 1 (O1): To explore wireless mechanisms for an adequate integration of semantic communications principles, ensuring support for time-sensitive applications with low latency and low jitter in multi-user, mobile environments. The aim is to explore the fundaments of Shannon theory and to study network coding approaches, as well as how to best suit them to ensure deterministic guarantees such as bounded latency (1-2 ms). This is done while maximizing the number of endpoints being served.
• Objective 2 (O2): To explore context-aware, decentralized learning approaches to best support a concept of IIoT SemCom. SemComIIoT shall focus on the study of decentralized learning approaches across Edge-Cloud to best support the required encoding/decoding of SemCom.
• Objective 3 (O3): To provide a holistic design for the integration of Semantic Communications in IIoT scenarios. The aim is to, based on the studies developed, propose a semantic communications architecture, capable of achieving end-to-end resilience and supporting deterministic behaviour of the infrastructure, whenever required. Having in mind a converged network service for different traffic types including time-critical control and high-definition video.

Project supported by the Sino-German Center (SGC), jointly supported by Deutsche Forschungsgemeinschaft (DFG) and National Natural Science Foundation of China (NSFC).
Grant agreement nr: M-0626

01.01.2023 - 31.12.2024