Toward Natively Intelligent Semantic Communications and Networking

As the process of envisioning, researching, and strategizing for 6G communications has commenced, we present our vision on the role of semantic communications (SemCom) in the beyond 5G era. In the following discourse, we delve into the factors that may impede the progress of research and implementation of 6G with the objective of catering to the semantically enriched communication requirements of the 2030s. In addition, we proceed to establish the fundamental attributes of SemCom and engage in an examination of the requisite technological modalities. In order to bolster this overarching vision, we introduce a semantic networking architecture that encapsulates and characterizes various application scenarios of SemCom. Expanding on the concept of point-to-point SemCom systems the semantic networking architecture focuses on extracting and filtering goal specific semantic information at the source before transmitting signals. It also involves decoding and post processing semantics at the destination. What sets this networking architecture apart is its transformation into a multi-user distributed, edge-to-cloud network with deadline constraint traffic. Given its complexity, implementing such an architecture necessitates frameworks for extracting and representing knowledge theoretical models to predict and manage multiple time varying deadline/delay constrained traffic flows that may cause significant congestion in the network as well, as innovative metrics infused with semantics to measure performance while encapsulating its inherent relevance. Finally, we provide research considerations and future directions towards the integration of semantics in the forthcoming 6G wireless systems.