Introduction
As 5G and future wireless generations strive for fiber-like data rates, the integration of Millimeter Wave (MMW) and Free-Space Optical (FSO) technologies becomes essential. FSO offers massive bandwidth but is vulnerable to atmospheric conditions like fog, while MMW-based RF links provide high speeds with better resilience to weather. While point-to-point hybrid links are well-studied, real-world deployments often involve complex multi-hop "daisy chains" or mesh networks. This paper provides a comprehensive analytical framework for evaluating these sophisticated topologies, specifically accounting for non-ideal hardware and advanced error-correction protocols like Hybrid Automatic Repeat Request (HARQ).
Methodology
The researchers developed a mathematical framework to predict outage probability and data rates across mixed-media networks. They employed a Rician fading model for MMW-RF links and Exponential or Gamma-Gamma distributions for FSO links to account for atmospheric turbulence. A key differentiator of this work is the integration of a state-of-the-art model for Power Amplifier (PA) efficiency, acknowledging real-world hardware impairments.
To handle the mathematical complexity of multi-hop mesh networks, the authors developed several approximation paths based on codeword length:
- Central Limit Theorem (CLT): Applied to "long codewords" where many channel realizations occur during a single transmission.
- Minkowski Inequality/Bounding: Applied to "short codewords" where the channel remains relatively static.
The framework's validity was confirmed through extensive Monte Carlo simulations, ensuring that the "near-closed-form" expressions accurately reflect performance across various Signal-to-Noise Ratio (SNR) regimes.
Key Insights
The "Mapping" Phenomenon
A significant finding is the existence of mathematical mappings between RF and FSO hops. By scaling specific channel parameters, an RF hop can be modeled to "look" like an FSO hop in terms of outage probability. This simplifies the analysis of heterogeneous networks by allowing them to be treated under a unified mathematical umbrella.
HARQ as a Hardware Compensator
The study demonstrates that HARQ is not merely an error-correction tool but a powerful mechanism to mitigate hardware impairments. In networks with inefficient PAs, HARQ effectively "hides" these shortcomings, bringing performance levels close to those of ideal hardware.
Antenna Sensitivity and Mesh Benefits
While network outage probability becomes insensitive to the number of RF antennas beyond a small threshold (e.g., $N \ge 3$), the ergodic rate (average data rate) continues to scale significantly with additional antennas. Furthermore, mesh networks provide a massive boost through path diversity; for a target outage of $10^{-6}$, moving from a single route to a mesh with multiple parallel paths can reduce the required SNR at each hop by approximately 1.2 dB.
Results
The analytical approximations (Lemmas 1-6) match numerical simulations almost perfectly, providing a reliable shortcut for network design. Key performance metrics include:
- Energy Efficiency: For a standard dual-hop setup with a target outage of $10^{-4}$, implementing HARQ with three retransmissions reduces the required transmit power by 17 dB compared to open-loop systems.
- Rate Matching: The paper introduces a specific formula (Corollary 2) to calculate the minimum number of RF antennas required to prevent the RF segment from becoming a bottleneck for high-speed FSO links.
Conclusion and Future Work
While this framework offers robust tools for designing hybrid MMW-FSO networks, some limitations remain. The current mesh analysis assumes non-overlapping routes, potentially underestimating inter-link interference in dense deployments. Future research could expand this to include Amplify-and-Forward relaying strategies and account for weather correlation across multiple hops, where a single atmospheric event might simultaneously degrade several FSO links.
Summary updated