STAR-RIS辅助ISAC系统波束赋形优化方法

Research on Beamforming Optimization of STAR-RIS Assisted ISAC System

本文提出了将通信感知一体化(Integrated Sensing and Communication,ISAC)系统与同时透射和反射可重构智能表面(Simultaneously Transmitting and Reflecting Reconfigurable Intelligent Surfaces,STAR-RIS)相结合的系统模型,解决了传统可重构智能表面RIS不能实现全空间通信与感知的问题;同时考虑在STAR-RIS上安装低成本的专用传感器以实现ISAC系统在STAR-RIS上执行目标感知的一种新颖有源架构,解决了雷达感知的严重路径损耗问题.本文旨在联合优化ISAC基站处的波束赋形和STAR-RIS的无源波束赋形,以最大化通信用户的加权和速率(Weighted Sum Rate,WSR),同时保证感知性能的最低信噪比(signal-to-noise ratio,SNR).为了解决该复杂非凸优化问题,交替优化基站波束赋形及STAR-RIS无源波束赋形.针对所提的满足雷达感知SNR最低要求下最大化WSR问题,基站处波束赋形的优化子问题等价为加权最小均方误差(Weighted Minimum Mean Square Error,WMMSE)问题,STAR-RIS处无源波束赋形优化子问题等价为分式规划(Fractional Programming,FP)问题.进一步,分别将优化的非凸子问题转化为二次约束二次规划(Quadratic Constraint Quadratic Programming,QCQP),并使用半正定松弛(Semidefinite Relaxation,SDR)技术将它们分别转化为凸的半正定规划(Semidefinite Programming,SDP)子问题进行迭代求解.仿真结果验证了所采用新型STAR-RIS辅助ISAC方案的优点和所提算法在提高WSR性能上的有效性.

This paper proposes a system model that combines the integrated sensing and communication(ISAC)system with simultaneously transmitting and reflecting reconfigurable intelligent surfaces(STAR-RIS),solving the problem that traditional reconfigurable intelligent surface RIS cannot achieve full space communication and perception;Simultaneously considering the installation of low-cost specialized sensors on STAR-RIS to achieve a novel active architecture for ISAC system to perform target perception on STAR-RIS,solving the serious path loss problem of radar sensing.This article aims to jointly optimize beamforming at ISAC base stations and passive beamforming at STAR-RIS to maximize the weighted sum rate(WSR)of communication users while ensuring the lowest signal-to-noise ratio(SNR)of sensing performance.To solve this complex non convex optimization problem,alternate optimization of base station beamforming and STAR-RIS passive beamforming is carried out.The optimization subproblem for beamforming at the base station is equivalent to the minimum weighted mean square error(WMMSE)problem for maximizing WSR under the minimum requirement of radar perception SNR.The optimization subproblem for passive beamforming at STAR-RIS is transformed into a fractional programming(FP)problem.Furthermore,the optimized non convex subproblems are respectively transformed into quadratic constraint Quadratic programming(QCQP),and they are respectively transformed into convex semi definite programming(SDP)subproblems for iterative solution using semi definite relaxation(SDR)technology.The simulation results validate the advantages of the new STAR-RIS assisted ISAC scheme and the effectiveness of the proposed algorithm in improving WSR performance.