Design and Optimization of THz Integrated Sensing and Communication Systems

A Terahertz (THz) Full-Duplex (FD) Integrated Sensing and Communication (ISAC) system is investigated in this paper, where an array-of-subarray (AoSA) architecture, with each radio-frequency (RF) chain dedicated to an individual subarray, is adopted. In the temporal domain, a frame is systematically divided into distinct segments for sensing and communication. Notably, this architecture allows for differentiated power allocations between these two functional segments, satisfying the corresponding performance requirements. Then, simultaneous time and power optimization problems with the objective of enhancing sensing and communication subject to the maximum transmitted power and sensing or rate constraints are formulated Subsequently, the maximum sensing range is determined under these dual constraints. A critical threshold of the sensing range is identified, enabling the optimal use of maximum power and a single time slot for sensing to achieve superior communication rate performance while guaranteeing the sensing requirements. In order to address these optimization challenges, the optimization algorithm is designed, namely the Gradient Descent (GD)-based algorithm. Numerical results are provided to justify the theoretical characterizations and the performance improvements of the proposed algorithm. Jointly optimizing the power and time allocation results in significant improvements in system performance and an expanded sensing range, on the basis of lower algorithm complexity. Also, a frequency dependent sensing range is identified in the THz band. Overall, the impact of operation frequency and joint power and time allocation design on the THz FD ISAC system is characterized.