Overhead- and energy-aware integrated sensing and communication for IoT: A dual-mode edge-intelligent framework

Abstract

The Integrated Sensing and Communication (ISAC) is an important direction for Internet of Things (IoT) networks because it uses the same radio infrastructure for data delivery and environmental awareness through shared spectrum, pilots, and hardware. There are many IoT-specific challenges that have not yet been resolved. Specifically, the assumptions regarding sensing overhead, latency, memory burden, and energy feasibility of passive nodes have not been adequately considered in prior ISAC formulations. This paper proposed an overhead and energy-aware ISAC framework that switches between passive, active, and hybrid sensing modes based on traffic load, scene dynamics, and harvested energy feasibility. A simulation campaign of 1500 Monte Carlo episodes, between the proposed framework and dedicated sensing plus communication, and static ISAC baselines, was performed. OEA-ISAC achieved the highest system utility (41.10 ± 0.12), the highest sensing F1-score (0.937 ± 0.001), the lowest latency (22.32 ± 0.20 ms), and the lowest energy per delivered Mbit (0.536 ± 0.003 J/Mb). OEA-ISAC has a 19.9% utility improvement over static ISAC, while it has a 27.3% utility improvement over dedicated sensing plus communication. The findings suggest that including sensing overhead as a primary control variable is a valuable area of future research in the context of hybrid active/passive IoT systems.