Sum Rate optimization for RIS-Aided RSMA system with Movable Antenna

This paper proposes a movable-antenna-assisted Rate-Splitting Multiple Access (RSMA) framework integrated with a Reconfigurable Intelligent Surface (RIS) to maximize sum rate by jointly optimizing beamforming, reflection coefficients, rate splitting, and antenna positions, demonstrating significant performance gains over conventional fixed-antenna systems.

The Gist

Imagine you are trying to host a massive dinner party in a house with very strange, cluttered furniture. You want to talk to every guest at the same time without your voices overlapping and becoming a confusing mess. This is exactly the challenge modern wireless networks face: sending data to many phones simultaneously without the signals crashing into each other.

This paper proposes a brilliant new way to solve this problem by combining three high-tech "superpowers": RSMA, RIS, and Movable Antennas.

Here is the breakdown of their solution using simple analogies:

1. The Problem: The "Cluttered Room"

In a normal wireless system, the signal travels from the tower (Base Station) to your phone. But if there are walls, buildings, or trees in the way, the signal gets blocked or bounces around chaotically.

• The Old Way: The tower has fixed antennas stuck in one spot. It's like a waiter standing in one corner of the room trying to shout instructions to everyone. If a guest moves behind a pillar, they can't hear.
• The Interference: When the waiter shouts to Guest A, Guest B might hear it too and get confused.

2. The Three Superpowers

A. RSMA (Rate-Splitting Multiple Access): The "Smart Waiter"

Instead of shouting a single message to everyone, the waiter (the tower) splits every message into two parts:

1. The Common Part: A general announcement everyone needs to hear (like "Dinner is served!").
2. The Private Part: Specific instructions just for one person (like "John, your steak is rare").

The waiter shouts the "Common Part" loudly so everyone hears it. Once everyone has heard the general announcement, they ignore it and focus only on their "Private Part." This prevents the messages from stepping on each other's toes. It's a very flexible way to manage noise.

B. RIS (Reconfigurable Intelligent Surface): The "Magic Mirror Wall"

Imagine the room has a wall covered in thousands of tiny, smart mirrors.

• How it works: If the waiter's voice is blocked by a pillar, the "Magic Mirror Wall" catches the sound and bounces it perfectly around the corner to the guest.
• The Benefit: It creates a clear path where there used to be a wall. The mirrors can be adjusted instantly to focus the signal exactly where it's needed.

C. Movable Antennas (MA): The "Dancing Waiter"

This is the paper's big innovation.

• The Old Way: The waiter is tied to a post.
• The New Way: The waiter is on wheels! They can slide around the room on a track.
• The Benefit: If the waiter notices that Guest A is in a "dead zone" (a spot where the signal is weak), the waiter can physically slide a few feet to the left or right to find the perfect spot to shout clearly. This gives the system a whole new dimension of freedom to dodge obstacles.

3. The Big Idea: Putting It All Together

The authors asked: "What happens if we combine the Smart Waiter (RSMA), the Magic Mirror Wall (RIS), and the Dancing Waiter (Movable Antennas)?"

They realized that while the Magic Mirror and the Smart Waiter are great, the Dancing Waiter adds a massive boost. By moving the antennas to the perfect spot while the mirrors adjust and the messages are split, the system becomes incredibly efficient.

4. How They Solved It (The "Recipe")

Optimizing all these things at once is like trying to solve a Rubik's cube while juggling. You can't just move one piece; moving the waiter changes how the mirrors need to work, which changes how the messages should be split.

The authors created a step-by-step algorithm (a recipe) to solve this:

1. Split the message: Figure out the best way to divide the "Common" and "Private" parts.
2. Adjust the mirrors: Tell the RIS how to bounce the signals.
3. Move the waiter: Slide the antennas to the best physical location.
4. Repeat: Do this over and over, getting slightly better each time, until the system is perfectly tuned.

5. The Result: A Huge Win

When they tested this in a computer simulation, the results were impressive:

• Compared to a system with fixed antennas, adding the "Dancing Waiter" (Movable Antennas) boosted the total speed (Sum Rate) by about 33% to 35%.
• It worked even better with the "Smart Waiter" (RSMA) than with older methods.

The Bottom Line

This paper shows that the future of 6G internet isn't just about sending stronger signals; it's about making the hardware flexible. By letting antennas move like dancers, using smart mirrors to bounce signals, and splitting messages intelligently, we can create wireless networks that are faster, clearer, and much harder to block by obstacles.

In short: They figured out how to make the radio tower "dance" around obstacles to deliver your data faster than ever before.

Technical Summary

Here is a detailed technical summary of the paper "Sum Rate optimization for RIS-Aided RSMA system with Movable Antenna".

1. Problem Statement

The paper addresses the performance limitations of conventional Rate-Splitting Multiple Access (RSMA) systems combined with Reconfigurable Intelligent Surfaces (RIS) when using Fixed-Position Antennas (FPAs). While RSMA offers robust interference management and RIS enhances signal propagation by shaping the wireless environment, fixed antennas restrict the system's spatial degrees of freedom (DoF), limiting the ability to fully exploit channel conditions, especially in non-line-of-sight (NLOS) scenarios.

The authors propose a novel framework integrating Movable Antennas (MAs) into a downlink Multi-Input Single-Output (MISO) RSMA-RIS system. The core objective is to maximize the system sum rate by jointly optimizing four coupled variables:

1. Transmit Beamforming Matrix ($W$): Managing signal direction and power allocation.
2. RIS Reflection Matrix ($\Phi$): Controlling phase shifts to optimize the reflected path.
3. Common-Rate Partition ($r_{c,k}$): Determining how much of the common message rate is allocated to each user.
4. Movable Antenna Positions ($x$): Dynamically adjusting the physical location of antenna elements within a defined region to optimize channel gains.

The optimization problem is highly non-convex due to the coupling of these variables, the non-linear nature of the achievable rate functions, and constraints such as constant modulus for RIS elements and minimum spacing between movable antennas.

2. Methodology

To solve the complex joint optimization problem, the authors propose an efficient Alternating Optimization (AO) algorithm. The methodology involves the following key steps:

• Problem Reformulation via Fractional Programming (FP):
  The original non-convex objective function and constraints are transformed into a more tractable form using Fractional Programming. This introduces auxiliary variables ($\mu, \epsilon, \gamma, v$) to decouple the numerator and denominator of the signal-to-interference-plus-noise ratio (SINR) expressions.

• Closed-Form Common Rate Allocation:
  The authors derive a closed-form solution for the common rate partition. They prove that the optimal strategy is an equal distribution of the total achievable common rate among all users, specifically $r^*_{c,k} = \frac{1}{K} \min_k \{R_{c,k}\}$. This simplifies the problem by reducing the number of variables to optimize iteratively.

• Iterative Subproblem Solving:
  The main problem is decomposed into four subproblems, solved sequentially in each iteration:

  1. Beamforming Matrix ($W$): Updated using Karush–Kuhn–Tucker (KKT) conditions. The authors derive closed-form update rules for the beamforming vectors based on Lagrange multipliers, avoiding high-complexity convex solvers like CVX.
  2. RIS Reflection Matrix ($\Phi$): Solved via a dual ascent method. The problem is converted into a dual problem where Lagrange multipliers are updated to satisfy constraints, and the RIS phase shifts are updated using a gradient-based approach to maximize the lower bound of the objective function.
  3. Movable Antenna Positions ($x$): Updated using a Gradient Ascent method. The gradient of the objective function with respect to antenna positions is calculated, and positions are iteratively adjusted. A step-size reduction mechanism is employed to ensure the updated positions remain within the feasible region (satisfying boundary and minimum spacing constraints).
  4. Auxiliary Variables: Updated via closed-form expressions derived from the FP transformation.

• Algorithm Convergence:
  The proposed algorithm (Algorithm 1) iterates through these updates until the sum rate converges or a maximum iteration count is reached.

3. Key Contributions

• First Integration of MA and RSMA-RIS: To the best of the authors' knowledge, this is the first work to investigate the performance gains of Movable Antennas specifically within an RSMA-RIS system.
• Joint Optimization Framework: The paper formulates a comprehensive sum-rate maximization problem that simultaneously optimizes beamforming, RIS reflection, rate splitting, and physical antenna positioning.
• Efficient Algorithm Design: The authors developed a low-complexity AO algorithm. Key innovations include the closed-form solution for rate splitting, KKT-based beamforming updates, and dual-problem-based RIS optimization, which collectively reduce computational burden compared to brute-force methods.
• Theoretical Validation: The use of KKT conditions and dual ascent ensures that the solution adheres to the necessary optimality conditions for the non-convex problem.

4. Numerical Results

The performance of the proposed P-MA-RSMA scheme was evaluated against several benchmarks, including FPA-based RSMA/SDMA and other MA-based schemes (CFGS-based and ORIS-based).

• Convergence: The proposed algorithm converges rapidly, typically stabilizing within 10 iterations.
• Performance Gains:
  • MA vs. FPA: Integrating MAs into RSMA-RIS systems yields significant gains. At a power budget of 46 dBm, MA provides a 35.6% performance enhancement for RSMA compared to FPA-based RSMA. For SDMA, the gain is approximately 33.3%.
  • RSMA vs. SDMA: RSMA consistently outperforms SDMA, particularly as the power budget increases. The sum rate of MA-RSMA increased by 81.8% when power rose from 34 dBm to 46 dBm, compared to a 64.1% increase for MA-SDMA. This highlights RSMA's superior degrees of freedom.
  • User Scalability: As the number of users ($K$) increases from 2 to 12, the sum rate of MA-RSMA grows by 47.1%, significantly outperforming FPA-RSMA (29.2% growth).
• Comparison with Benchmarks: The proposed algorithm outperforms existing MA-RSMA schemes (C-MA-RSMA and O-MA-RSMA) across various power budgets and user counts.

5. Significance

This work demonstrates that physical layer flexibility (via Movable Antennas) combined with signal processing flexibility (via RSMA) and environmental reconfigurability (via RIS) creates a synergistic effect that significantly boosts 6G network capacity.

• Overcoming Blockages: The system is particularly effective in NLOS scenarios where direct links are blocked, as the combination of RIS and MA can dynamically reshape the channel to find optimal propagation paths.
• Hardware Efficiency: By allowing antennas to move, the system achieves high performance without requiring a massive increase in the number of fixed antenna elements, offering a cost-effective solution for future dense networks.
• Future Direction: The study validates that joint optimization of spatial (position) and signal (beamforming/rate) domains is crucial for next-generation wireless systems, setting a precedent for future research in 6G architecture.