Non-orthogonal Multiple Access for Multi-cell Indoor VLC

Uday Thummaluri, Abhinav Kumar, Lakshmi Natarajan

Abstract

In this letter, we propose a non-orthogonal multiple access (NOMA) based scheme for multi-cell indoor visible light communications (VLC), where cell-edge user is jointly served by multiple cells. Unlike the typical designs in the existing literature, the proposed scheme doesn't involve complex domain and direct current bias addition, which makes the design simple and feasible for real time implementation in indoor VLC. The symbol error rate (SER) of the proposed NOMA scheme is analysed using analytical and simulation results, and is compared with orthogonal multiple access (OMA). It is observed that the average SER of the users with the proposed NOMA is marginally degraded as compared to OMA. However, the SER of cell-edge user with the proposed NOMA is significantly improved with joint maximum likelihood decoding, and outperforms successive interference cancellation based decoding and OMA, with trade-off on computational complexity.

Introduction: Addressing Interference in Indoor VLC

Visible Light Communication (VLC) leverages indoor LED lighting to provide high-speed data networking. A significant challenge in multi-cell VLC environments is the performance of users at the cell edges. These "cell-edge users" often suffer from severe interference from adjacent light fixtures, which degrades signal reliability.

Traditional solutions frequently employ NOMA in conjunction with Optical OFDM (e.g., DCO-OFDM). However, these methods typically require Hermitian symmetry to generate real-valued signals for LED transmission, which halves the effective data rate. Additionally, they necessitate a Direct Current (DC) bias to maintain signal positivity, leading to power inefficiencies and increased hardware complexity. This research proposes a simplified NOMA scheme that operates entirely in the real domain, specifically designed to enhance the reliability of cell-edge users without these overheads.

Proposed NOMA System Model

The proposed power-domain NOMA scheme is modeled for a two-cell, three-user indoor scenario. In this setup, two transmitters ($Tx_1$ and $Tx_2$) serve three users: $U_1$ (near $Tx_1$), $U_3$ (near $Tx_2$), and $U_2$, the cell-edge user positioned to receive signals from both transmitters.

System model and channel parameters for the multi-cell indoor VLC setup.

Real-Domain Encoding and Joint Transmission

The scheme maps data directly to constellation points (power levels) in the real domain. This avoids the computational complexity of Inverse Fast Fourier Transforms (IFFT) and the data rate loss associated with Hermitian symmetry. Crucially, the cell-edge user ($U_2$) is served jointly by both transmitters, effectively turning potential interference into a useful signal component.

Decoding Strategies

Different decoding strategies are employed based on user location:

  • Center Users ($U_1, U_3$): These users utilize Successive Interference Cancellation (SIC). They first decode and subtract the stronger signal intended for the edge user before decoding their own data.
  • Edge User ($U_2$): This user employs Joint Maximum Likelihood (JML) decoding. JML evaluates all possible combinations of transmitted signals from both cells to identify the most likely transmitted symbols, providing a significant diversity gain over standard SIC.

Key Results and Performance Analysis

The performance of the proposed scheme was evaluated using numerical simulations based on standard Lambertian radiation patterns and VLC channel models, including parameters such as a 4m room height and a 60° LED semi-angle.

SER performance for individual users, comparing simulation and analytical results.

Edge User Superiority and Analytical Validation

As shown in the performance plots, the Symbol Error Rate (SER) for the cell-edge user ($U_2$) is significantly improved when using JML decoding compared to OMA or SIC-based NOMA. The simulation results closely match the derived analytical closed-form expressions across a range of Transmit SNR (110 dB to 160 dB).

While the average SER across all users shows a marginal degradation compared to OMA, this is offset by the substantial reliability gains for the most vulnerable user ($U_2$). This trade-off is central to the proposed design: prioritizing the performance of edge users who would otherwise experience poor connectivity.

Computational Complexity

The improved performance of JML decoding comes with a trade-off in computational complexity. JML requires searching a larger space of constellation combinations compared to SIC, making it more computationally intensive. However, for the center users, the complexity remains low, and the overall system avoids the high-speed processing overhead required by OFDM-based subcarrier management.

Conclusion and Future Directions

The proposed NOMA scheme offers a hardware-friendly alternative for multi-cell indoor VLC by operating in the real domain and utilizing joint transmission for edge users. While it significantly boosts edge-user reliability, future work is needed to address scalability in larger environments where JML complexity might become prohibitive. Additionally, investigating performance under imperfect Channel State Information (CSI) and developing adaptive power allocation strategies remain important areas for further research.

Summary updated

Abstract

In this letter, we propose a non-orthogonal multiple access (NOMA) based scheme for multi-cell indoor visible light communications (VLC), where cell-edge user is jointly served by multiple cells. Unlike the typical designs in the existing literature, the proposed scheme doesn't involve complex domain and direct current bias addition, which makes the design simple and feasible for real time implementation in indoor VLC. The symbol error rate (SER) of the proposed NOMA scheme is analysed using analytical and simulation results, and is compared with orthogonal multiple access (OMA). It is observed that the average SER of the users with the proposed NOMA is marginally degraded as compared to OMA. However, the SER of cell-edge user with the proposed NOMA is significantly improved with joint maximum likelihood decoding, and outperforms successive interference cancellation based decoding and OMA, with trade-off on computational complexity.
BibTeX
@article{thummaluri2022non,
  author = {Uday Thummaluri and Abhinav Kumar and Lakshmi Natarajan},
  title = {{Non-orthogonal Multiple Access for Multi-cell Indoor VLC}},
  journal = {IEEE Wireless Communications Letters},
  year = {2022},
  url = {https://openreview.net/forum?id=T5XtwG-LIn},
}
Figure