Transceiver Optimization for Multi-Hop Communications With Per-Antenna Power Constraints

In multiple-input multiple-output (MIMO) systems, power constraints are critically important for transceiver designs. The most popular and widely used power constraint is sum power constraint. However, each antenna usually has its own amplifier, per-antenna power constraints are a more accurate and reasonable power constraint model. In this paper, under per-antenna power constraints a framework is proposed for the transceiver designs for multi-antenna multi-hop cooperative communications. In the framework, both linear transceiver and nonlinear transceiver designs with various objective functions are investigated. In specific, the considered nonlinear transceiver designs include transceivers with decision feedback equalizer (DFE) or Tomlinson-Harashima precoding (THP). To find the optimal solutions, we tackle the problem both numerically and analytically. In the proposed framework, based the matrix-monotonic optimization framework, the optimal architectures of the optimal solutions are derived, and the relationship between the designs under sum power constraint and per-antenna constraints is clearly revealed. It is proved that per-antenna power constraints are in nature equivalent to a weighted sum power constraint. After introducing a weighting matrix for the power constraints, closed-form optimal solutions can be derived. Finally, simulation results assess the performance of the proposed transceiver designs.