Inverse Search: An ISAC Resource Allocation Algorithm Searching Inward from the Outside of the Pareto Boundary
Tianshu Jiang, Li Chen, Pengcheng Zhu · 2025
Integrated Sensing and Communication (ISAC) combines sensing and communication functions into a unified system, thereby enhancing the overall performance of the network and enabling a broader range of application scenarios, including autonomous driving, the industrial internet, and intelligent healthcare. ISAC now is recognized as one of the key technologies for 6 G. The uncertainty of channel conditions and the growing demand for customized services from users present challenges to the resource allocation of ISAC. Most existing allocation schemes begin their search at the lower performance bound within the feasible domain. The value of this lower bound significantly influences the trajectory of the search route. When there is a substantial angle between the search route and the Pareto boundary, it can lead to unstable system performance. In this article, we analyze the issue of performance decline resulting from user heterogeneity and unstable channel gains. Then, we propose an inverse search scheme named Weighted Chebyshev Compromise (WCC) method. Our proposed scheme conducts a search from the expected performance outside the feasible domain, ensuring that the search direction is approximately perpendicular to the Pareto boundary, aided by channel gain. Finally, through simulation, we demonstrated that this approach effectively mitigates performance degradation caused by user heterogeneity and unstable channel gains.