⚡ electrical engineering

Unlocking Downlink NOMA with FARIS: Joint Clustering and Surface Configuration Design

This paper proposes a two-stage optimization framework that jointly designs user clustering, power allocation, and fluid active reconfigurable intelligent surface (FARIS) configurations to maximize the sum rate of a downlink NOMA system, demonstrating near-optimal performance and superior efficiency compared to existing benchmarks.

Original authors: Hong-Bae Jeon, Tuo Wu

Published 2026-08-18

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Hong-Bae Jeon, Tuo Wu

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 a world where the air itself can be tuned like a radio dial, turning a chaotic, unpredictable wireless signal into a clear, directed stream of information. For decades, engineers have tried to tame the wireless environment, which naturally scatters and weakens signals as they travel from a tower to a phone. A major breakthrough in this field came with the invention of "intelligent surfaces"—walls or panels covered in tiny, smart elements that can bounce signals around obstacles to reach a user. However, these early versions had a significant flaw: they were passive. Like a mirror reflecting a dim candle, they could only bounce what they received, often making the signal even weaker as it traveled through the air twice. To fix this, researchers developed "active" surfaces that could amplify the signal, but these were stuck in a fixed shape, unable to move their elements to find the best path. Another innovation allowed elements to move fluidly to find better spots, but these moving parts lacked the power to boost the signal. The challenge remained: how to combine the ability to move with the ability to amplify, all while serving many users at once without them stepping on each other's signals.

This paper introduces a solution that merges these capabilities into a single, powerful system called a "fluid active reconfigurable intelligent surface," or FARIS. The researchers focused on a specific method of sending data called non-orthogonal multiple access, or NOMA. In simple terms, NOMA allows a base station to send messages to multiple users simultaneously by layering them on top of one another, much like stacking different voices on a single radio frequency. For this to work, the system needs to create a clear difference in signal strength between users so that the receiver can peel away the layers one by one. The problem is that in a normal, static environment, these differences are often too small or unpredictable. The authors propose that by using a FARIS, the system can dynamically choose exactly where to place its reflecting elements and how strongly to amplify the signal, effectively sculpting the air to create the perfect conditions for this layered transmission.

To test this idea, the team built a mathematical model of a network where a base station talks to dozens of users through this smart, moving, amplifying surface. They faced a massive puzzle: they needed to decide which users should be grouped together, how much power to give each user, which specific elements on the surface should be turned on, where those elements should be positioned, and how to adjust their phase to steer the signal. Doing all of this at once is incredibly complex, with millions of possible combinations. The researchers developed a two-step strategy to solve it. First, they grouped users based on their distance from the surface, ensuring that each group contained people with naturally different signal strengths, which makes the layering process easier. Then, for each group, they used a sophisticated, step-by-step computer algorithm to fine-tune the surface's position, amplification, and phase settings. This algorithm works by adjusting one setting at a time while holding the others steady, repeating the process until the system settles on the best possible configuration.

The results of their simulations were striking. When they compared their new FARIS system against existing technologies, it consistently delivered much higher data speeds. In their tests, the new system outperformed standard active surfaces that cannot move, and also beat moving surfaces that cannot amplify. The advantage was most visible when the system had to serve many users at once or when the available power was limited. The researchers found that by jointly optimizing the movement and the amplification, the system could overcome the natural weakening of signals that usually plagues wireless networks. They also verified that their computer method was highly efficient, finding solutions that were nearly as good as the absolute best possible answer, but in a fraction of the time. The study confirms that combining fluid mobility with active amplification creates a new design space that significantly boosts the speed and reliability of wireless connections, offering a promising path forward for the next generation of communication networks.

Technical Summary: Unlocking Downlink NOMA with FARIS

Problem Statement
This paper addresses the performance limitations inherent in current Reconfigurable Intelligent Surface (RIS)-assisted Non-Orthogonal Multiple Access (NOMA) systems. Conventional passive RIS (PRIS) suffers from severe multiplicative fading due to cascaded path loss, while Active RIS (ARIS) mitigates this attenuation but remains constrained by a fixed, static geometry that limits spatial adaptability. Conversely, Fluid RIS (FRIS) offers spatial reconfigurability but lacks the active amplification capability to overcome signal attenuation. The paper identifies a critical gap: existing solutions treat signal amplification and spatial adaptability as separate problems. Furthermore, NOMA performance relies heavily on sufficient channel disparity among users to enable Successive Interference Cancellation (SIC), a condition often difficult to guarantee in static propagation environments. The authors propose investigating a Fluid Active Reconfigurable Intelligent Surface (FARIS)-aided downlink NOMA system, where each element possesses both fluid mobility and active amplification, to jointly optimize spatial locations and signal reinforcement.

Methodology
The authors formulate a network-wide sum-rate maximization problem that jointly optimizes user clustering, NOMA power allocation, FARIS amplification gains, discrete phase shifts, and fluid element selection. The problem is subject to Quality-of-Service (QoS) constraints, reflected-power limits, and hardware power consumption constraints. Due to the nonconvex and mixed-integer nature of the problem (MINLP), the authors develop a two-stage framework:

Distance-Based Interleaved Clustering: A lightweight strategy is employed to construct NOMA user groups. Users are sorted by large-scale channel gains (distance) and assigned to clusters in an interleaved manner. This approach ensures inherent channel disparity within clusters, facilitating effective SIC with negligible overhead.
Per-Cluster Alternating Optimization (AO): For each cluster, the remaining variables are optimized using an AO framework that decomposes the problem into four tractable subproblems:
- Power Allocation: Solved using Geometric Programming (GP) with arithmetic-geometric mean approximation (AGMA) to handle fractional constraints.
- Amplification Gain Design: Addressed via Fractional Programming (FP) using Lagrangian dual transforms and quadratic transforms, followed by a Majorization-Minimization (MM) approach to handle nonconvex constraints.
- Discrete Phase Shift Optimization: The continuous phase problem is relaxed and solved using MM with quadratic minorization. The discrete constraint is then handled by formulating a Mixed-Integer Linear Program (MILP) using one-hot binary variables.
- Fluid Element Selection: A Cross-Entropy Method (CEM) is utilized to efficiently search the combinatorial space of element subsets, iteratively updating a probability distribution based on elite samples.

Key Contributions

System Modeling: The paper establishes a new system model for FARIS-aided downlink NOMA. This model captures a coupled spatial-amplitude signal structure where the FARIS simultaneously determines the spatial positions of reflecting elements, their active reflection gains, and discrete phase shifts. It explicitly accounts for amplifier noise propagation and hardware power consumption associated with fluid-active ports.
Problem Formulation: A comprehensive optimization problem is formulated that integrates user clustering, power allocation, and FARIS configuration under realistic constraints, including discrete phase quantization and binary element selection.
Algorithmic Framework: The proposed two-stage framework effectively decouples the combinatorial clustering from the continuous/discrete optimization, providing a tractable solution to the highly nonconvex MINLP.
Performance Validation: Extensive simulations validate the framework, demonstrating rapid convergence and near-optimal performance relative to brute-force search (BFS).

Results
Numerical results indicate that the proposed FARIS-NOMA system consistently outperforms benchmark schemes, including FARIS-OMA (Orthogonal Multiple Access), FRIS-NOMA (passive fluid), and ARIS-NOMA (active static).

Sum-Rate Gains: The proposed scheme achieves a 16.5% to 30.3% increase in network sum-rate compared to baselines at typical transmit power levels.
Convergence: The AO framework exhibits rapid and stable convergence, with the objective value increasing monotonically.
Optimality: Compared to BFS in small-scale settings, the proposed algorithm achieves approximately 94.6% to 95.2% of the optimal sum-rate.
Parameter Sensitivity: The system demonstrates robust performance across varying numbers of selected elements ( $N_o$ ), candidate elements ( $N$ ), phase resolution ( $b$ ), and cluster sizes ( $K_t$ ). Notably, the joint exploitation of fluid reconfiguration and active reflection allows the system to effectively manage intra-cluster interference and convert power budget into rate gains more efficiently than static or passive counterparts.

Significance
The paper claims that the integration of FARIS and NOMA unlocks a fundamentally new design space by jointly optimizing where the signal is reflected (spatial mobility) and how strongly it is reinforced (active amplification). This dual capability allows the system to induce favorable channel disparities for efficient SIC while compensating for cascaded path loss. The authors posit that this approach effectively addresses the limitations of both multiplicative fading and restricted spatial degrees-of-freedom (DoF) found in conventional architectures. The study concludes that the proposed framework offers a practical and high-performance solution for smart radio environments in 6G, verifying that the joint exploitation of fluid reconfiguration, active reflection, and power-domain multiplexing yields substantial gains in network throughput.

More like this