arXiv:2503.01383

Channel Semantic Characterization for Integrated Sensing and Communication Scenarios: From Measurements to Modeling

Zhengyu Zhang, Ruisi He, Bo Ai, Mi Yang, Xuejian Zhang, Ziyi Qi, Zhangdui Zhong

제출 2025-03-03

원문 초록

With the advancement of sixth-generation (6G) wireless communication systems, integrated sensing and communication (ISAC) is crucial for perceiving and interacting with the environment via electromagnetic propagation, termed channel semantics, to support tasks like decision-making. However, channel models focusing on physical characteristics face challenges in representing semantics embedded in the channel, thereby limiting the evaluation of ISAC systems. To tackle this, we present a novel framework for channel modeling from the conceptual event perspective. By leveraging a multi-level semantic structure and characterized knowledge libraries, the framework decomposes complex channel characteristics into extensible semantic characterization, thereby better capturing the relationship between environment and channel, and enabling more flexible adjustments of channel models for different events without requiring a complete reset. Specifically, we define channel semantics on three levels: status semantics, behavior semantics, and event semantics, corresponding to channel multipaths, channel time-varying trajectories, and channel topology, respectively. Taking realistic vehicular ISAC scenarios as an example, we perform semantic clustering, characterizing status semantics via multipath statistical distributions, modeling behavior semantics using Markov chains for time variation, and representing event semantics through a co-occurrence matrix. Results show the model accurately generates channels while capturing rich semantic information. Moreover, its generalization supports customized semantics.

함께 살펴볼 논문

공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.

토론

이 논문과 연결된 포럼 글이 아직 없습니다.