Joint Spectrum Partitioning and Power Allocation for Energy Efficient Semi-Integrated Sensing and Communications

TL;DR

This work tackles spectrum congestion by introducing a semi-ISaC framework that supports communication-only, sensing-only, and ISaC services. It develops joint spectrum partitioning and power allocation schemes to maximize a combined sensing information and data-rate metric, and separately to maximize energy efficiency, accounting for service priorities, clutter, QoS constraints, and power limits. The MI and data-rate objective is shown to be convex, while the EE objective is tackled as a concave-convex fractional program solved via a Dinkelbach-based iteration, yielding global optimality within the convex subproblems. Simulations demonstrate substantial gains over benchmarks, reveal how resource allocation shifts with QoS and priority settings, and confirm fast convergence, underscoring practical benefits for energy-efficient, multi-service wireless networks.

Abstract

With spectrum resources becoming congested and the emergence of sensing-enabled wireless applications, conventional resource allocation methods need a revamp to support communications-only, sensing-only, and integrated sensing and communication (ISaC) services together. In this letter, we propose two joint spectrum partitioning (SP) and power allocation (PA) schemes to maximize the aggregate sensing and communication performance as well as corresponding energy efficiency (EE) of a semi-ISaC system that supports all three services in a unified manner. The proposed framework captures the priority of the distinct services, impact of target clutters, power budget and bandwidth constraints, and sensing and communication quality-of-service (QoS) requirements. We reveal that the former problem is jointly convex and the latter is a non-convex problem that can be solved optimally by exploiting fractional and parametric programming techniques. Numerical results verify the effectiveness of proposed schemes and extract novel insights related to the impact of the priority and QoS requirements of distinct services on the performance of semi-ISaC networks.

Figures (7)

• Figure 1: Considered semi-ISaC system with sensing-only, ISaC, and communications-only services.
• Figure 2: Spectrum Partition as a function of sensing MI threshold ($R_r$) and data rate threshold ($R_c$) when $R_c=R_r$.
• Figure 3: Aggregate sensing MI and data rate as a function of sensing MI threshold ($R_r$) and data rate threshold ($R_c$).
• Figure 4: Sensing MI or data rate of each user as a function of the priority of ISaC user and minimum QoS requirements.
• Figure 5: Aggregate sensing MI and data rate as a function of transmit power of Semi-ISaC BS and target RCS.