Finite-Alphabet-Aware Trajectory and Precoder Optimization for UAV Relaying

Haoyang Di, Xiaofeng Zhu, Yulin Shao2025 · DOI

TL;DR

This paper introduces a novel framework for optimizing Unmanned Aerial Vehicle (UAV) relaying systems that accounts for the practical reality of finite-alphabet signals, moving beyond idealized Gaussian assumptions. By jointly optimizing the UAV's flight path and its transmission strategy, the proposed method achieves significant performance gains, demonstrating the critical need to align communication system design with real-world constraints.

What Problem Does This Paper Address?

Modern wireless communication systems, especially those employing UAVs as relays, often rely on simplified models that assume transmitted signals follow a Gaussian distribution. While this assumption simplifies mathematical analysis and optimization, it deviates from how actual systems operate, which use discrete, finite-alphabet constellations (like BPSK, QPSK, etc.). This mismatch can lead to suboptimal designs for UAV trajectories and transmission precoders, potentially wasting energy and reducing data throughput in real-world deployments. The paper tackles this gap by developing a framework that explicitly incorporates the constraints of finite-alphabet signals into the joint optimization of UAV trajectory and precoder design.

Methods

The core of the paper's contribution lies in its novel optimization framework. The problem of jointly optimizing the UAV's trajectory and precoders for finite-alphabet inputs is inherently non-convex and complex due to the coupling between these variables and the structure of the mutual information for discrete signals. To address this, the authors propose an efficient solution based on Alternating Optimization (AO) and Successive Convex Approximation (SCA).

The AO approach decomposes the complex joint optimization problem into two sub-problems:

  1. UAV Trajectory Optimization: Given fixed precoders, the UAV's flight path is optimized.
  2. Precoder Optimization: With a fixed UAV trajectory, the precoding matrices at the base station and the UAV are optimized.

For each sub-problem, SCA is employed to handle the non-convexity. SCA works by approximating the non-convex functions with convex ones at each iteration, allowing standard convex optimization solvers (like CVX) to find a local optimum. This iterative process, alternating between optimizing the trajectory and the precoders, is repeated until convergence, yielding a globally optimal solution for the formulated problem.

Datasets

Not discussed in the paper. The evaluation relies on simulated data generated with specific parameters.

Key Insights and Ideas

  • The "Gaussian Illusion": The paper powerfully demonstrates that designs optimized under the assumption of Gaussian signals can be significantly suboptimal in practice. This highlights a critical disconnect between theoretical modeling and real-world system performance.
  • Finite-Alphabet Matters: Explicitly modeling finite-alphabet inputs is not just an academic exercise; it leads to tangible improvements in system performance, including better throughput and potentially energy efficiency.
  • Joint Optimization is Crucial: The paper emphasizes that optimizing the UAV's trajectory and precoders jointly is essential for maximizing performance under practical constraints. Separating these optimizations would likely lead to suboptimal outcomes.
  • AO + SCA Synergy: The combination of Alternating Optimization and Successive Convex Approximation provides a robust and tractable method for solving highly complex, non-convex problems that arise from realistic system modeling.

Key Results

The paper presents simulation results that compare their proposed Finite-Alphabet-Aware Optimization (FAAO) framework against other schemes (MMSE-AO, MRT-AO, ZF-AO) that are optimized under Gaussian assumptions.

  • Performance Gains: The FAAO framework consistently outperforms the Gaussian-based schemes, showcasing its effectiveness in practical scenarios.
  • Trajectory Adaptability: Figure 2 illustrates how the UAV's trajectory adapts based on the available time resources. With limited time, the UAV takes a direct path. With more time, it exhibits more complex paths, potentially hovering or adjusting its position to optimize channel conditions before reaching its final destination.
  • Information Rate Improvement: Figure 3 shows that the FAAO framework achieves higher information rates compared to other schemes, especially as the transmit power (W) increases. This underscores the advantage of accounting for finite-alphabet signals.

Limitations and Future Work

  • Computational Complexity: While the proposed AO+SCA algorithm is efficient and tractable, solving complex optimization problems iteratively can still be computationally intensive, especially for large-scale systems.
  • Local Optima: Like many optimization techniques for non-convex problems, the AO+SCA approach guarantees convergence to a local optimum, not necessarily the global optimum.
  • Future Work: The authors suggest that their work provides a foundation for further research in realistic UAV-enabled communications. Potential future directions could include exploring more complex modulation schemes, incorporating additional practical constraints (e.g., interference, energy harvesting), and investigating dynamic environments.

Summary updated

BibTeX
@misc{di2025finite,
  author = {Haoyang Di and Xiaofeng Zhu and Yulin Shao},
  title = {{Finite-Alphabet-Aware Trajectory and Precoder Optimization for UAV Relaying}},
  year = {2025},
  doi = {10.1109/globecom59602.2025.11432294},
}
Figure