TL;DR
This paper tackles the challenge of optimizing both the light transmission (beamforming) and receiver orientation (PD orientation) in mobile Visible Light Communication (VLC) systems. They propose novel algorithms for both fixed and random user orientations, demonstrating significant improvements in Quality of Service (QoS) and reduced power consumption compared to existing methods.
What Problem Does This Paper Address?
Visible Light Communication (VLC) offers a promising alternative to traditional radio-frequency communication, leveraging existing LED infrastructure for high-speed, license-free data transmission. However, VLC is highly sensitive to line-of-sight (LOS) propagation, making it susceptible to blockages, especially from the user's own body or movement. This can lead to severe performance degradation and even communication outages. To combat this, researchers are exploring "Orientation Adjustable Receivers" (OARs) that can dynamically adjust their photo-detector (PD) orientation to better align with the incoming light.
The core problem this paper addresses is how to jointly optimize two critical aspects of a mobile VLC system:
- Transmit Beamforming (BO): How the LEDs transmit light to maximize signal strength and coverage.
- PD Orientation (PD): How the receiver's PD is oriented to capture the incoming light effectively.
This joint optimization is crucial because these two aspects are coupled and influence each other. Furthermore, the mobility of the user (UE) introduces complexities, as their orientation can change, leading to imperfect channel conditions and requiring robust solutions. Existing work often assumes perfect alignment or doesn't fully consider the interplay between beamforming and receiver orientation, especially in dynamic scenarios.
Methods
The authors propose a two-pronged approach based on the user's orientation:
- Fixed UE Orientation:
- They first model the OAR and human body blockage.
- The joint optimization problem of beamforming and PD orientation is formulated but found to be non-convex and intractable due to the coupling.
- To overcome this, they decompose the problem into two sub-problems: a beamforming subproblem and a PD orientation subproblem.
- An Alternating Optimization (AO) algorithm is developed to iteratively solve these sub-problems, progressively refining both the transmit beamformer and the PD orientation.
- For the beamforming subproblem, they employ Semidefinite Relaxation (SDR) to handle non-convexity and convert it into a convex Semidefinite Program (SDP).
- For the PD orientation subproblem, they use an iterative optimization and projection method.
- Random UE Orientation:
- They introduce a model for random UE orientation and derive an imperfect Channel State Information (CSI) model.
- A robust joint BO optimization problem is formulated to minimize transmit power while ensuring the worst-case Quality of Service (QoS).
- This problem is also tackled using an AO approach, but with added robustness to handle the imperfect CSI.
- The S-Procedure is utilized to convert infinite linear matrix inequality constraints arising from the robust formulation into finite ones.
Datasets
Not discussed in the paper. The evaluation is performed through numerical experiments based on the proposed system models.
Key Insights and Ideas
- Joint Optimization is Key: The paper strongly emphasizes that optimizing beamforming and PD orientation together is essential for maximizing performance in mobile VLC, rather than treating them independently.
- OARs Offer Significant Gains: The introduction of Orientation Adjustable Receivers (OARs) provides a substantial advantage by allowing the receiver to adapt to changing light conditions, leading to lower transmit power requirements and improved QoS.
- Mobility Demands Robustness: For mobile users, simply optimizing for ideal conditions is insufficient. Robust optimization techniques are necessary to guarantee performance under varying and imperfect channel states caused by random user orientation.
- Decomposition and Alternating Optimization: The non-convexity of the joint optimization problem is effectively managed by decomposing it into smaller, solvable sub-problems and iteratively optimizing them using an AO framework.
- SDR for Beamforming: Semidefinite Relaxation proves to be a powerful tool for transforming the non-convex beamforming subproblem into a tractable convex SDP.
Key Results
- The proposed joint BO optimization schemes (for both fixed and random UE orientation) significantly enhance the Quality of Service (QoS) of mobile VLC users.
- These schemes achieve this enhancement with reduced transmit power consumption compared to benchmark schemes (e.g., non-OAR systems).
- The AO algorithm for fixed UE orientation converges within a few iterations, demonstrating its efficiency.
- The robust design for random UE orientation effectively handles angle errors and ensures the minimum rate requirement, showing less conservatism than robust designs without OAR.
- Specifically, Figure 8(a) shows that transmit power decreases with increasing elevation angle threshold for OAR, and is generally lower than non-OAR schemes. Figure 8(b) shows transmit power increases with rate threshold, with OAR performing better. Figure 9(a) shows achievable rate R varies periodically with user location, with OAR consistently outperforming non-OAR. Figure 10 shows the robust OAR design guarantees the minimum rate requirement more reliably than non-robust schemes.
Limitations and Future Work
- Computational Complexity: While the proposed algorithms are effective, the computational complexity can still be significant, especially for the robust optimization in random UE orientation scenarios. The paper mentions the complexity of the AO algorithm for problem (22) as approximately $O(\max \{N+2, 4\}^4(N+3)^{1/2} \log(1/\zeta_1))$, where $\zeta_1$ is the solution accuracy.
- Idealized Models: The paper relies on specific models for OAR, human blockage, and channel gain. The performance in real-world scenarios might differ if these models are not perfectly representative.
- Future Work: The paper implicitly suggests future work in exploring more complex scenarios, such as multi-user interference, different types of blockages, and more advanced OAR technologies. The robustness of the proposed algorithms to various levels of CSI uncertainty is also a key area for further investigation.
Summary updated