TL;DR
This paper proposes a novel hybrid transmission strategy combining TDMA, NOMA, and cooperative transmission for multi-UAV collaborative sensing in cell-free MIMO networks. By jointly optimizing task allocation and transmission strategies, the authors significantly reduce mission completion time, outperforming existing methods.
What Problem Does This Paper Address?
The increasing demand for real-time aerial data collection, such as surveillance, environmental monitoring, and disaster response, necessitates efficient data transmission from multiple Unmanned Aerial Vehicles (UAVs) to a central cloud server (CS). Traditional approaches often struggle with the challenges of inter-UAV interference, limited bandwidth, and the need to minimize overall mission completion time. Existing research has explored various aspects like UAV trajectory optimization and single-UAV communication, but a comprehensive solution for multi-UAV collaborative sensing in advanced network architectures like cell-free MIMO, which offers enhanced spectral efficiency and coverage, has been lacking. This paper aims to bridge this gap by developing an efficient strategy that leverages the strengths of different multiple access techniques to accelerate data collection from multiple UAVs.
Methods
The core of this work lies in a novel hybrid transmission strategy that integrates three key techniques:
- Time-Division Multiple Access (TDMA): Used for private sensing data transmission, ensuring interference-free communication for each UAV's unique data.
- Non-Orthogonal Multiple Access (NOMA): Employed for a portion of the private sensing data, allowing multiple UAVs to transmit simultaneously on the same resource, managed by successive interference cancellation (SIC) at the CS.
- Cooperative Transmission: Utilized for the common sensing task, where all UAVs contribute to transmitting shared data, leveraging the collective capabilities of the network.
To optimize this hybrid approach, the authors formulate a joint optimization problem that considers:
- Task Splitting: Determining the optimal ratio of common sensing tasks versus individual private sensing tasks for each UAV.
- Transmission Strategy Optimization: Dynamically allocating resources and choosing between TDMA and NOMA for private data, and optimizing cooperative transmission for common data.
The non-convex nature of this optimization problem is addressed by reformulating it using fractional programming (FP) and matrix FP techniques. An alternating optimization algorithm is then proposed, which iteratively optimizes different sets of variables to achieve a monotonically decreasing mission completion time, leading to a locally optimal solution.
Datasets
Not discussed in the paper. The paper relies on simulated data for numerical evaluation.
Key Insights and Ideas
- Synergy of Multiple Access Techniques: The paper demonstrates that a hybrid approach combining TDMA, NOMA, and cooperative transmission is more effective than using any single technique alone. Each technique addresses specific challenges: TDMA for interference-free private data, NOMA for efficient resource utilization, and cooperative transmission for shared data.
- Task Splitting is Crucial: The optimal division of sensing tasks between common and private components significantly impacts mission completion time. The proposed algorithm dynamically determines these splits based on system conditions.
- Cell-Free MIMO Advantage: The cell-free MIMO architecture, with its distributed access points (APs), plays a vital role in enabling effective cooperative reception and interference management, which is crucial for the success of the proposed hybrid strategy.
- NOMA's Role in Interference Management: While TDMA inherently avoids interference, NOMA, when combined with advanced decoding at the CS (like SIC), can outperform TDMA by allowing simultaneous transmissions and better utilizing the available resources, especially in cell-free environments.
Key Results
The paper presents extensive numerical results validating the effectiveness of the proposed hybrid TDMA-NOMA-cooperative transmission strategy. Key findings include:
- Significant Reduction in Mission Completion Time: The proposed scheme consistently outperforms baseline schemes (TDMA only, NOMA only, cooperative only, and hybrid with equal task allocation) across various scenarios. For instance, in Figure 4, the proposed scheme shows a substantial reduction in completion time compared to others as SNR increases.
- Rapid Convergence: The alternating optimization algorithm converges quickly, typically within a few iterations, as demonstrated in Figure 3.
- Performance Gains with Increased UAVs and APs: Increasing the number of UAVs (K) and APs (M) generally improves performance for all schemes, but the proposed hybrid strategy consistently maintains its advantage (Figures 7 and 8).
- Impact of Data Size and SNR: As the amount of sensory data (btotal) increases, cooperative transmission becomes more advantageous, and the proposed scheme adapts accordingly (Figure 5). High SNR levels also favor cooperative transmission.
- Task Splitting Enhances Performance: Figure 9 clearly shows that task splitting significantly accelerates sensing and transmission processes for both TDMA and NOMA, and the proposed hybrid scheme with optimized task splitting achieves the shortest mission completion time.
- Altitude Impact: While increasing UAV altitude generally increases mission completion time due to pathloss, the cooperative and proposed hybrid schemes exhibit the smallest increase, demonstrating their robustness (Figure 10).
Limitations and Future Work
- Non-Globally Optimal Solution: The proposed alternating optimization algorithm guarantees a locally optimal solution due to the inherent non-convexity of the problem. The authors suggest running the algorithm with multiple random initial points to increase the probability of finding a global solution.
- Fixed SIC Decoding Order: The optimization of the SIC decoding order in the NOMA phase is left for future study.
- UAV Placement Optimization: The paper acknowledges that optimizing UAV trajectories and placements could yield further gains but leaves this as a future research direction due to the complexity involved in modeling pathloss and NLoS components.
- CSI Imperfections: The robustness of the system against imperfect channel state information (CSI) is not explored and is suggested for future work.
- Comparison with Other Techniques: A comprehensive comparison with other multiple access techniques like OFDMA and RSMA is suggested for future research.
Summary updated