Robust Downlink Transmission Design in IRS-Assisted Cognitive Satellite and Terrestrial Networks
Bai Zhao,Min Lin,Ming Cheng,Jun-Bo Wang,Julian Cheng,Mohamed-Slim Alouini
DOI: https://doi.org/10.1109/jsac.2023.3288234
IF: 16.4
2023-01-01
IEEE Journal on Selected Areas in Communications
Abstract:Cognitive satellite and terrestrial network (CSTN) is considered as a promising technology to provide ubiquitous connectivity for various users within wide-coverage. This paper proposes a robust downlink transmission scheme for multiple intelligent reflecting surfaces (IRSs) assisted CSTN. Here, the satellite network adopts multigroup multicast transmission scheme to serve many earth stations, while the terrestrial network exploits space division multiple access and multi-IRS-enhanced non-orthogonal multiple access technology to communicate with many terrestrial users. By assuming that these two networks share the same frequency band having only the angular information based imperfect channel state information of each user, we formulate an optimization problem to minimize the total transmit power subject to the constraints of quality-of-service requirement for each user, per-antenna transmit power budgets of satellite and BS, and unit-modulus requirement for each reflecting element. To tackle this mathematically intractable problem, we then employ angular discretization together with the successive convex approximation method to obtain the active beamforming (BF) vectors of satellite and BS, the passive BF vector of IRS, and the power allocation coefficients. Moreover, we propose a generalized zero forcing BF and alternative optimization to obtain the suboptimal solutions of the optimization problem with low computational complexity. Finally, simulation results are given to demonstrate the effectiveness and superiority of the proposed two schemes over the benchmarks.
telecommunications,engineering, electrical & electronic