TL;DR
This paper proposes a novel approach to enhance communication capacity in millimeter-wave (mmWave) networks by employing a full-duplex unmanned aerial vehicle (FD-UAV) as a relay. The core contribution is a joint optimization framework for the UAV's position, its beamforming capabilities, and its power control strategy to maximize end-to-end data rates, outperforming existing methods.
What Problem Does This Paper Address?
The paper tackles the challenge of limited coverage and high path loss in mmWave communication systems, especially when direct line-of-sight (LoS) links are obstructed by ground-based obstacles. While UAVs offer a promising solution due to their mobility and ability to establish LoS links, integrating them into mmWave networks introduces complexities. These include mitigating self-interference (SI) in full-duplex systems, optimizing the UAV's placement for maximum signal strength, and efficiently managing power for both the UAV and ground nodes. Existing solutions often address these issues in isolation, leading to suboptimal overall performance.
Methods
The authors propose a three-pronged approach:
- Conditional Optimal Positioning: They first derive a closed-form solution for the UAV's optimal position under ideal conditions (LoS environment and perfect beamforming). This provides a strong initial estimate for the UAV's location. The UAV is then deployed to the closest feasible position that guarantees LoS links.
- Alternating Interference Suppression (AIS) Algorithm: For joint optimization of beamforming vectors (BFVs) and power control, they introduce the AIS algorithm. This iterative method alternates between optimizing BFVs to maximize target signal gains and minimize interference, and then optimizing power control variables.
- Power Control Optimization: A theorem is presented to determine the optimal transmit powers for the source node (SN) and the FD-UAV relay to maximize the minimum achievable rate between the SN and the destination node (DN), while avoiding power waste.
Datasets
Not discussed in the paper. The paper relies on simulated data generated based on established channel models and specified simulation parameters.
Key Insights and Ideas
- Joint Optimization is Crucial: The paper strongly emphasizes that optimizing UAV positioning, beamforming, and power control jointly is essential for achieving near-optimal performance in mmWave FD-UAV relay systems.
- Ideal Beamforming as a Starting Point: Using ideal beamforming to derive the initial UAV position significantly simplifies the problem and provides a robust starting point for the iterative optimization.
- AIS Algorithm's Effectiveness: The proposed AIS algorithm effectively handles the non-convexity and high dimensionality of the problem, leading to rapid convergence and near-upper-bound performance.
- Robustness to Beam Misalignment: The proposed method demonstrates robustness to beam misalignment, with achievable rates remaining close to the upper bound even with moderate deviations in angles.
- Trade-off in Interference Suppression: There's a clear trade-off between aggressive interference suppression and maximizing target signal gains, which the AIS algorithm implicitly manages.
Key Results
- The proposed joint positioning, beamforming, and power control scheme significantly outperforms three benchmark schemes ("RandPos & AIS", "DesPos & steer", and "DesPos & OMP").
- The method closely approaches a performance upper bound for mmWave FD-UAV relay systems, achieving rates within 0.3 bps/Hz for 4x4 antenna arrays and even closer for 8x8 arrays.
- The AIS algorithm converges quickly, typically within 4 iterations, and its convergence speed is influenced by the step size parameter κ.
- As transmit power increases, the achievable rate also increases, but with diminishing returns, especially when the UAV is positioned optimally relative to the DN or SN.
- The performance is sensitive to the SN-DN distance, with achievable rates decreasing as the distance increases due to path loss.
Limitations and Future Work
- Ideal Beamforming Assumption: The initial positioning step relies on an ideal beamforming assumption, which might not always hold in real-world scenarios. The paper addresses this by searching for a feasible LoS position if the initial ideal position doesn't guarantee LoS.
- Computational Complexity: While the AIS algorithm is efficient, the overall computational complexity is still dependent on the search space for positioning and the number of iterations.
- Future Work: The paper suggests studying the convergence of Algorithm 1 more thoroughly. Additionally, exploring the impact of hardware impairments and more complex channel models could be future research directions. The robustness to beam misalignment is demonstrated, but further investigation into advanced beam tracking mechanisms could be beneficial.
Summary updated