A Context-Aware Semantic Resource Allocation Framework for Integrated Sensing and Communication in Vehicular Networks

Abstract

The intersection of the Integrated Sensing and Communication (ISAC) and semantic communication is an op portunity for change in the next generation of vehicular networks. Nonetheless, current strategies do not have the capability of dynamically optimizing the resource allocation by collectively addressing dynamic wireless channel conditions, differentiating heterogeneous data semantics, and dynamic vehicular conditions. Thus, this paper suggests a new framework, Context-Aware Semantic Resource Allocation (CASRA), that posits resource allocation in the context of a multi-objective optimization prob lem, which seeks to maximize contextual semantic utility by realistically looking at physical constraints. The framework presents a dynamic weighting system, which calculates the semantic importance, depending on the criticality of messages, vehicle condition, and network properties. The full simulation in realistic conditions of 3GPP vehicular scenarios proves CASRA yields a semantic fidelity of 0.897, an increase of 66.7% over the Channel-Aware allocation, 41.4% over SA-VA-ISAC, and 76.5% over NOMA-based Semi-ISAC. Moreover, it minimizes the emergency communication latency by 35.8% over Channel Aware allocation with a mean latency of 12.14 ms, and it achieves resource efficiency of 31.8% over the next-best approach. Under congested, emergency, and highway conditions, CASRA is highly performing and efficient and equitable at the same time. These findings suggest that CASRA offers a promising approach toward reliable, efficient, and intelligent 6G vehicular ecosystems.

Keywords