COMSEMA
COMSEMA
ANR PROJECT 2024-2028
Embedded Files
Semantic Communications
The principle is to take into account the nature of the data, and to transmit only the information that is relevant to the purpose of the communication.
A semantic communication system typically comprises the following key components:
The semantic encoder component is responsible for detecting and extracting the semantic content from the source message. Furthermore, it may compress or eliminate irrelevant information, thus enhancing efficiency.
The function of the channel encoder is to encode and modulate the semantic features of the message as signals to mitigate any noise or interference that may occur during transmission.
Upon reception of the signal, the channel decoder demodulates and decodes it, thereby recovering the transmitted semantic features.
The semantic decoder interprets the semantic features into a format that is comprehensible to the destination user.
A knowledge base provides the semantic encoder and decoder with a foundation upon which they can accurately and effectively understand and infer semantic information.
Overview
Wireless networks are currently witnessing a radical shift from a purely data-oriented architecture to service and intelligent-based architectures, allowing hence the support of a diverse set of verticals. Thanks to the development of AI, future networks are expected to incorporate an even larger set of applications and services such as ReID applications and human activity recognition, interactive hologram, e-health, intelligent humanoid robot, etc. To avoid the huge amount of data transmitted by these services over the wireless networks, the fundamental semantics-blind approach to communication, so far prevalent in today’s wireless systems, should be questioned. In fact, thanks to the development of generative AI and LLM-based techniques, it is nowadays possible to extract the semantic representations of an application and to transmit them instead of exchanging the whole data. However, when transmitted over wireless networks, these representations/embeddings are subject to interference, noise, and channel fadings.
In this project, we consider video interpretation applications and propose a fundamental semantics-approach to redesign the entire process of information generation and transmission in the network. More specifically, we will provide a comprehensive study on the impact of wireless errors on the semantic mismatch/errors and design novel semantic encoding/decoding schemes, that consider the disruptions of the wireless network, going thus beyond current methods that assume an error-free environment between the semantic encoder and decoder. Furthermore, we propose a shift from current bit error-free transmission metrics towards generic tasks related metrics and develop novel wireless transceivers that compress further the data transmitted over the wireless system. Finally, novel AI-based interference management that focuses on the task achievement, rather than the bit rate improvement over the air interface, will be investigated.
Workpackages
WP1: Use Cases, KPIs and Dissemination
T1.1 – Scenarios, KPIs and Evaluation methodology
T1.2 – Monitoring semantic existing work and dissemination activities
WP2: AI-based semantic extraction for video interpretation applications
T2.1 – Enhancing construction of semantic representation over wireless networks
T2.2 – Efficient encoding digital features for network of camera
WP3: Semantic-based communication design and optimization
T3.1 – Semantic-based transceiver design
T3.2 – Energy-efficient interference management
Publications
@ In submission
Semantic Communication based on Matryoshka Representation Learning and Product Quantization
Authors : Leonardo Roque, Quentin Lampin, Louis-Adrien Dufrène, Guillaume Larue, Grégoire Lefebvre and Mohammad Assaad
Disseminations
@ IEEE Meditcom 2025 - Workshop on Emerging Technologies in Smart Semantic Communications for 6G
Semantic Communications. Keynote presented by Louis-Adrien Dufrène
Authors : Quentin Lampin, Louis-Adrien Dufrène and Guillaume Larue
Contacts
Centrale Supélec – Laboratoire L2S
Orange Research
INRIA – Teams STARS
Page updated
Google Sites
Report abuse