⚡ electrical engineering

Outage Probability Analysis of MRC-Based Fluid Antenna Systems under Rician Fading

This paper analyzes the outage probability of a fluid antenna system employing maximum ratio combining under Rician fading by deriving exact and asymptotic statistics for both physical and virtual reference port models, revealing that the virtual model provides superior accuracy, especially as system size and signal quality increase.

Original authors: Tummi Ganesh, Soumya P. Dash, Italo Atzeni

Published 2026-07-07

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Tummi Ganesh, Soumya P. Dash, Italo Atzeni

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to catch a radio signal in a busy city. The signal is bouncing off buildings, cars, and trees, making it wobble and fade in and out. This is what engineers call "fading."

To fix this, traditional radios use multiple antennas (like having several ears) to listen to the signal from different angles. But in the future (6G), we want to pack even more antennas into tiny devices, which is hard to do with fixed, rigid metal rods.

The Solution: The "Liquid" Antenna
This paper introduces a new idea called a Fluid Antenna System (FAS). Imagine a single, flexible antenna that can physically move or "flow" to different spots within a small area. Think of it like a single microphone that can instantly slide to 10 different positions on a table to find the clearest spot to hear a speaker.

The paper studies a specific setup:

The Setup: A transmitter sends a signal to a receiver with this "fluid" antenna.
The Strategy: The fluid antenna has $M$ possible spots (ports) it can occupy. The receiver checks the signal strength at all $M$ spots, picks the best $K$ of them, and combines their signals together (like mixing the best audio tracks to make a louder, clearer song). This is called Maximum Ratio Combining (MRC).
The Goal: To calculate the "Outage Probability" (OP). In simple terms, this is the chance that the combined signal is still too weak to be understood, even after picking the best spots.

The Big Problem: The "Reference Port" Trap
When these antenna spots are very close together, they don't act independently; they "correlate." If one spot hears a bad signal, its neighbor likely hears a bad signal too.

To calculate how this correlation works, engineers usually use a math model. The paper compares two models:

Model A (The Old Way - Physical Reference Port): This model assumes the first spot on the antenna is the "boss" or "reference." It calculates how all other spots relate to that first one. The paper argues this is like trying to describe a family tree by only looking at how everyone relates to the oldest sibling. It's a simplification that gets messy and inaccurate when you have many spots.
Model B (The New Way - Virtual Reference Port): This model introduces a "ghost" or "virtual" reference point that doesn't physically exist but helps mathematically describe how all spots relate to each other equally. The paper claims this is the more accurate way to describe reality.

What the Paper Found
The authors did a deep mathematical dive to see how these two models affect the reliability of the connection.

The Math Works: They derived complex formulas to predict exactly how often the signal will fail (outage) for both models. They also found a "lower bound," which is a safe, easy-to-calculate estimate that is very close to the real answer.
The "Diversity" Bonus: They proved that if you have $M$ spots and pick the best $K$ , your ability to fight signal fading (diversity order) is equal to $M$ . Essentially, having more spots gives you a bigger safety net.
The Old Model is Flawed: This is the most important finding. The "Old Way" (Model A) looks okay when the signal is weak or when you only pick a few spots. However, as the signal gets stronger (high SNR) and as you use more spots ( $M$ $M$ and $K$ $K$ get bigger), the Old Way becomes wildly inaccurate.
- Analogy: Imagine trying to predict the weather. On a calm day, a simple guess might work. But during a massive storm with complex wind patterns, that simple guess fails completely. The paper shows the "Old Model" fails to predict the "storm" (high performance scenarios) correctly.
The "Concave" Surprise: When they tested how the signal behaves with different types of interference (Rician fading), the "Old Model" predicted a weird, upside-down "U" shape (performance gets worse before it gets better). The "New Model" showed the expected result: performance just keeps getting better as the signal gets clearer. This proves the Old Model is fundamentally misleading in certain conditions.

The Takeaway
If you are designing these next-generation fluid antennas, you cannot rely on the old, simplified math model that treats the first antenna as the boss. You must use the more accurate "Virtual Reference" model, especially when you have many antenna spots and are trying to get the best performance out of them. Ignoring this difference could lead to designing systems that think they are working perfectly when they are actually failing.

Technical Summary: Outage Probability Analysis of MRC-Based Fluid Antenna Systems under Rician Fading

Problem Statement
Fluid Antenna Systems (FAS) have emerged as a promising paradigm for 6G and beyond, offering spatial diversity without the hardware complexity of massive MIMO by dynamically switching a single radiating element among multiple ports. However, accurate performance analysis of FAS is complicated by the inevitable spatial correlation between closely spaced ports. While previous studies often utilized a "simplified channel model" that treats the first physical port as a reference to define correlations, this approach imposes structural restrictions on the correlation between arbitrary ports, potentially leading to inaccuracies. Furthermore, existing literature on FAS has predominantly focused on single-port activation under Rayleigh fading, leaving a gap in understanding the performance of multi-port selection (specifically selecting the best $K$ out of $M$ ports) under Rician fading conditions, which include Line-of-Sight (LoS) components common in practical 6G scenarios.

Methodology
The paper investigates a FAS where a single-antenna transmitter communicates with a receiver equipped with a one-dimensional fluid antenna containing $M$ ports. The receiver selects the best $K$ ports with the highest instantaneous Signal-to-Noise Ratios (SNRs) and combines them using Maximum Ratio Combining (MRC). The channel is modeled under Rician fading.

The core methodological contribution involves a comparative analysis of two distinct channel correlation models:

Physical Reference Port Model: The widely used approach where the first physical port serves as the reference, and correlations are derived relative to it (based on Jakes' model).
Virtual Reference Port Model: A more accurate approach where a virtual reference port is introduced to link all physical ports via a common correlation coefficient, avoiding the structural constraints of the physical reference model.

The authors derive the statistics of the post-combining SNRs using a characteristic function (c.f.) approach. This allows for the derivation of exact integral-form expressions for the Outage Probability (OP) for both models. To address the mathematical intractability of these exact integrals, the paper derives closed-form lower bounds for the OP and analyzes the asymptotic behavior at high average SNR.

Key Contributions

Dual-Model Analysis: The paper provides a rigorous derivation of OP expressions for MRC-based FAS under Rician fading using both the physical reference port model and the more accurate virtual reference port model.
Characteristic Function Approach: A c.f.-based framework is established to handle the statistical dependencies introduced by port selection and correlation, leading to exact OP expressions.
Closed-Form Bounds and Asymptotics: The study derives closed-form lower bounds for the OP and asymptotic expressions at high SNR. These results reveal that the diversity order of the considered MRC-based FAS is $M$ , regardless of the correlation model used.
Quantification of Model Inaccuracy: The paper numerically evaluates the discrepancy between the two models, demonstrating that the physical reference port model becomes increasingly inaccurate as the number of available ports ( $M$ ), the number of selected ports ( $K$ ), and the average SNR increase.

Results
Numerical results validate the analytical framework, showing that the derived expressions match Monte Carlo simulations. Key findings include:

Diversity Order: The system achieves a diversity order of $M$ , confirming the effectiveness of selecting and combining multiple ports.
Model Discrepancy: The ratio of the outage probabilities between the physical and virtual models ( $P_{out}^{p.r.p.} / P_{out}^{v.r.p.}$ ) deviates significantly from unity at moderate-to-high SNR. This deviation grows with larger $M$ and $K$ , and smaller inter-port spacing (smaller $W$ ).
Rician Factor Behavior: A critical finding is the difference in how the two models respond to the Rician factor ( $\kappa$ ). The virtual reference port model shows a monotonically decreasing OP as $\kappa$ increases (indicating improved performance with stronger LoS). In contrast, the physical reference port model exhibits a concave trend, where OP initially increases before decreasing, suggesting a misleading prediction of performance degradation at certain LoS strengths.
Port Selection: Selecting the best $K$ ports significantly outperforms random port selection, and performance improves as $K$ and $M$ increase.

Significance
The paper establishes that while the physical reference port model simplifies analysis, it introduces significant inaccuracies in predicting the performance of multi-port FAS, particularly in high-SNR regimes and when utilizing a large number of ports. The proposed virtual reference port model offers a mathematically tractable yet more accurate alternative for system design and performance evaluation. The work underscores the necessity of adopting accurate correlation models when analyzing FAS for 6G networks, especially as systems move toward continuous-aperture MIMO concepts involving dense port arrays. The findings suggest that relying on simplified models may lead to suboptimal system configurations or incorrect performance expectations in practical deployments involving LoS components.

More like this