TL;DR
This paper provides a comprehensive performance analysis of hybrid satellite-terrestrial networks enhanced by Intelligent Reflecting Surfaces (IRS). By comparing two configurations—mounting the IRS on a High-Altitude Platform (HAP) versus placing it on the ground—the authors demonstrate that aerial IRS nodes significantly outperform terrestrial ones, offering a robust path toward achieving the ubiquitous, high-speed connectivity promised by 6G.
What Problem Does This Paper Address?
As we move toward the 6G era, the goal is "ubiquitous connectivity"—meaning high-speed internet even in remote, infrastructure-deficient, or disaster-stricken areas. Traditional terrestrial networks can't reach everywhere, leading researchers to look toward Non-Terrestrial Networks (NTNs) involving satellites and High-Altitude Platforms (HAPs).
However, satellite-to-ground links face massive path loss and atmospheric interference. Intelligent Reflecting Surfaces (IRS)—arrays of small, software-controlled elements that reflect electromagnetic waves to "steer" signals—offer a solution to boost signal strength without the high power consumption of traditional relays. The core question this paper tackles is: Where should we put the IRS? Is it better to mount it on the HAP (Aerial IRS) or use it on the ground to reflect signals from a HAP relay (Terrestrial IRS)?
Methods
The authors develop and compare two distinct system models:
- System Model 1 (Aerial IRS): A satellite sends a signal to an IRS mounted on a HAP, which reflects it directly to a ground user.
- System Model 2 (Terrestrial IRS): A satellite sends a signal to a HAP acting as a decode-and-forward relay. The HAP then transmits to a ground-based IRS, which reflects the signal to the end user.
To make the analysis realistic, they use sophisticated statistical models for fading: Rician for line-of-sight satellite links, Shadowed Rician for the tricky transition between aerial and terrestrial environments, and Nakagami-$m$ for terrestrial paths.
The researchers derived exact, closed-form mathematical expressions for:
- Outage Probability: The likelihood the signal drops below a usable threshold.
- Ergodic Rate: The average maximum data rate the system can support.
- Average Symbol Error Rate (ASER): Performance across various complex modulation schemes (Square, Rectangular, Cross, and Hexagonal QAM).
Datasets
This is a theoretical and analytical study; therefore, no external datasets were used. The results were validated using extensive Monte-Carlo simulations performed in MATLAB and Mathematica to ensure the derived mathematical formulas matched simulated physical behavior.
Key Insights and Ideas
- The Aerial Advantage: The Aerial IRS configuration (SM-1) generally outperforms the Terrestrial IRS setup (SM-2). In the terrestrial model, the system's performance is often bottlenecked by the initial satellite-to-HAP relay link, whereas the aerial IRS acts as a powerful direct "signal mirror."
- The Power of Scale: Increasing the number of IRS elements ($N$) provides a massive boost. For example, in the aerial model, doubling the number of elements can increase the ergodic rate by approximately 2 bps/Hz.
- Modulation Matters: The paper highlights that Hexagonal QAM (HQAM) is superior to standard Square QAM for high-speed 6G applications. Because HQAM packs constellation points more efficiently, it achieves a lower error rate for the same amount of transmit power (roughly a 1 dB gain at 1024-QAM).
- Shadowing Sensitivity: The severity of "shadowing" (obstructions like trees or buildings) drastically shifts the power requirements. Moving from "heavy" to "light" shadowing can result in a transmit power gain of over 10 dB to maintain the same connection quality.
Key Results
- Transmit Power Gains: For a target outage probability of $10^{-8}$, a system with 150 IRS elements requires roughly 35 dBm less transmit power than a system with only 30 elements.
- Ergodic Rate: In the aerial IRS model, the system achieved rates ranging from 15 to 25 bps/Hz as the IRS array grew from 30 to 960 elements.
- Relay vs. IRS: The study confirms that while HAP relays are useful, the "passive" gain provided by a large IRS array can achieve similar or better performance with significantly lower energy costs.
Limitations and Future Work
The authors acknowledge a few simplifying assumptions that provide avenues for future research:
- Single Antennas: The current model assumes all nodes (satellite, HAP, user) have a single antenna. Real-world 6G systems will likely use Massive MIMO (multiple antennas).
- Perfect Phase Control: The analysis assumes the IRS can perfectly shift the phase of incoming waves. In practice, hardware limitations lead to "phase noise" or quantized phase shifts.
- Static Users: The study focuses on stationary users. Future work could incorporate high-mobility scenarios, such as providing connectivity to high-speed trains or aircraft.
Summary updated