Joint Optimization of Training and Precoder for Dual-Functional MIMO Systems

The evolution of communication systems shows a trend toward multifunctional integration, thus the joint design of training sequence and precoder for multifunctional purposes is of great significance. In this paper, we investigate the joint optimization of training sequence and precoder matrices for dual-functional multiple-input multiple-output (MIMO) systems under per-antenna power constraints, which considers the performance metrics of channel estimation, data transmission and target estimation simultaneously. A general fusion framework under per-antenna power constraints is established, where multiple linear constraints are transformed into a single weighted-sum constraint, and a modified subgradient algorithm is proposed to address it. Then, based on the fusion of positive semi-definite matrix-valued signal-to-noise ratios (SNRs), training sequence is optimized to strike a trade-off between channel estimation accuracy and sensing performance, and the optimal pilot based on the fusion structure is derived. The proposed algorithm solves mean square error minimization and mutual entropy maximization problems, achieving a balance between system performance and algorithm complexity. Based on the optimized training sequence, channel estimation error model is derived, and the corresponding precoder matrix is designed, which takes into account the performance of both data transmission and target estimation. Finally, numerical results are provided for demonstrating the performance of the proposed algorithms.