⚡ electrical engineering

Dual-Tier IRS-Assisted Mid-Band 6G Mobile Networks: Robust Beamforming and User Association

This paper proposes a robust dual-tier IRS-assisted framework for mid-band 6G networks that jointly optimizes beamforming and user association to overcome LoS blockages and support massive IoT connectivity with significantly lower complexity than exhaustive search methods.

Original authors: Muddasir Rahim, Soumaya Cherkaoui

Published 2026-02-03

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Muddasir Rahim, Soumaya Cherkaoui

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 send a message to a friend in a crowded, skyscraper-filled city. In the future (6G networks), we want to send these messages at incredibly high speeds using a specific "Golden Band" of radio waves (7–15 GHz). This band is great because it carries a lot of data, but it has a major flaw: it's easily blocked by buildings, trees, and other obstacles. If the direct path is blocked, the signal dies.

To fix this, the authors of this paper propose a clever two-layer system using "smart mirrors" to bounce the signal around obstacles.

The Problem: The "Golden Band" is Fragile

Think of the 6G "Golden Band" as a high-speed courier service. It's fast and can carry huge packages (data), but the couriers are very sensitive. If a building blocks their path, they can't get through. Previous solutions tried to build more couriers (small cell towers) everywhere, but that's expensive and energy-hungry. Another idea was to use mirrors on the ground (Terrestrial IRS) to bounce signals around corners. But in a dense city, a mirror on the ground might still be blocked by a tall building, leaving some people in "dead zones."

The Solution: A Two-Tier Mirror System

The authors suggest a dual-layer approach, like having mirrors on the ground and mirrors in the sky:

Ground Mirrors (TIRS): These are like standard mirrors mounted on buildings. They help bounce signals to people nearby.
Sky Mirrors (AIRS): These are mirrors mounted on low-flying drones or balloons (Low-Altitude Platform Stations). Because they are high up, they can see over buildings and reach people the ground mirrors can't.

By combining both, the network creates a flexible web of reflected signals that can navigate around almost any obstacle.

The Challenge: The "Seating Arrangement" Puzzle

Now, imagine you have 100 people (users) and 100 mirrors (IRSs). You also have a powerful transmitter (the Access Point) with many antennas. The goal is to get the highest total speed for everyone.

This creates a massive puzzle:

Which person should use which mirror?
How should the transmitter aim its beams to avoid people talking over each other?

If you try to solve this by checking every single possible combination (like trying every possible seating arrangement at a wedding), it would take a supercomputer centuries to finish. This is called an "exhaustive search," and it's too slow for real life.

The Paper's Approach: A Smart Matchmaker

Instead of checking every possibility, the authors created a smart, step-by-step strategy:

Cancel the Noise (Zero-Forcing): First, they use a technique called "Zero-Forcing." Imagine the transmitter is a conductor. Instead of letting everyone shout at once, the conductor directs the sound so that when it reaches a specific person, the noise from everyone else cancels out perfectly. This ensures the signal is clean.
The Stable Match: Next, they use a "matching algorithm." Think of this like a dating app or a school admissions process:
- Each person ranks the mirrors based on which one would give them the best speed.
- Each mirror ranks the people based on who gives it the best speed.
- They go through rounds of proposing and accepting. If a mirror gets a "better" proposal, it swaps partners.
- This continues until everyone is paired up in a way that is "stable"—meaning no one wants to switch because they are already with their best available option.

The Results: Almost Perfect, Much Faster

The authors tested this system using computer simulations. They compared their "Smart Matchmaker" method against:

The Exhaustive Search: The perfect but impossibly slow method.
Greedy Search: A method where people just grab the first available mirror without thinking ahead.
Random Search: Picking mirrors by flipping a coin.

The findings were impressive:

Their method achieved performance almost identical to the perfect (but slow) exhaustive search (within 2%).
It was significantly faster than the perfect method, making it practical for real use.
It beat the "Greedy" and "Random" methods by a wide margin (up to 44% faster in some cases).

Conclusion

In simple terms, this paper proves that by using a mix of ground and sky-based "smart mirrors" in the 6G "Golden Band," we can keep internet connections fast and reliable even in the most crowded, blocked-up cities. They solved the complex math problem of who connects to what by using a smart, step-by-step matching system that is nearly as good as the perfect solution but runs in a fraction of the time. This provides a blueprint for building the robust, high-speed 6G networks of the future.

Technical Summary: Dual-Tier IRS-Assisted Mid-Band 6G Mobile Networks

Problem Statement
The rapid expansion of Internet of Things (IoT) applications and data-intensive services (e.g., VR/AR) necessitates robust resource allocation in future sixth-generation (6G) networks. While millimeter-wave (mmWave) and Terahertz (THz) bands offer high bandwidth, they suffer from severe line-of-sight (LoS) blockages and high propagation losses, often requiring costly dense small-cell deployments. Consequently, the upper mid-band (UMB), specifically the 7–15 GHz range known as FR3, has emerged as the "Golden Band" for 6G due to its favorable balance between bandwidth availability and coverage. However, FR3 still faces significant propagation losses and frequent blockages in dense urban environments.

Existing solutions often rely on Ultra-Massive MIMO (UM-MIMO), which incurs high power consumption and hardware complexity, or terrestrial Intelligent Reconfigurable Surfaces (TIRS), which may leave coverage blind spots in obstructed environments. Furthermore, current research on IRS-assisted networks predominantly focuses on mmWave/THz bands and single-tier terrestrial deployments, leaving the integration of aerial IRS (AIRS) mounted on Low-Altitude Platform Stations (LAPS) with TIRS in the FR3 spectrum largely unexplored. The core challenge addressed in this paper is the Joint Beamforming and User Association (JBUA) problem in a dual-tier (TIRS + AIRS) FR3 network, formulated as a Mixed-Integer Nonlinear Program (MINLP) that is NP-hard to solve directly.

Methodology
The authors propose a novel framework combining TIRS and AIRS to ensure reliable connectivity under severe LoS blockages in the FR3 band. The system model consists of a single Access Point (AP) with $N$ antennas serving $K$ single-antenna users, assisted by $L$ IRSs (a mix of TIRSs and AIRSs). Direct AP-to-user links are assumed unavailable due to blockages.

To solve the NP-hard JBUA problem, the authors employ a problem decomposition strategy, breaking the original optimization into two sub-problems:

Beamforming Matrix Optimization: With the user-IRS association fixed, the authors utilize Zero-Forcing (ZF) beamforming at the AP. This linear precoding technique eliminates multi-user interference by using the pseudo-inverse of the cascaded channel matrix, thereby simplifying the Signal-to-Interference-and-Noise Ratio (SINR) calculation to depend primarily on noise.
User-IRS Association: With the beamforming matrix fixed, the association problem is modeled as a one-to-one matching problem. The authors propose a user-proposing deferred-acceptance algorithm (based on stable matching theory). In this algorithm, users rank IRSs based on expected data rates and propose to their most preferred IRS. IRSs tentatively accept the proposal yielding the highest data rate and reject others, iterating until a stable matching is reached.

The complexity of the proposed algorithm is proven to be polynomial, specifically $O(K \times L)$ , where $K$ is the number of users and $L$ is the number of IRSs.

Key Contributions

First Analysis of Dual-Tier IRS in FR3: To the authors' knowledge, this is the first study analyzing the impact of integrating both TIRS and AIRS specifically within the FR3 (7–15 GHz) band for 6G communications.
Formulation of JBUA as MINLP: The paper formulates the joint optimization of beamforming and user association to maximize the downlink sum rate, identifying it as a Mixed-Integer Nonlinear Program (MINLP) and proving its NP-hardness.
Efficient Decomposition and Solution: The authors decompose the complex problem into tractable sub-problems, solving them using ZF beamforming and a stable matching algorithm to construct a feasible suboptimal solution.
Performance Validation: The proposed method is shown to approach the performance of an exhaustive search (ES) baseline while maintaining significantly lower computational complexity.

Simulation Results
Comprehensive simulations were conducted using MATLAB with parameters typical for FR3 (15 GHz carrier, 400 MHz bandwidth) and a network featuring 256 AP antennas and IRSs with $100 \times 100$ elements. The proposed JBUA scheme was compared against three baselines: Exhaustive Search (ES), Greedy Search (GS), and Random Search (RS).

Sum Rate vs. Transmit Power: The proposed JBUA achieved a sum rate within 2% of the optimal Exhaustive Search benchmark. It consistently outperformed the GS and RS baselines across the entire power range, yielding approximately 14% and 44% higher sum rates, respectively.
Sum Rate vs. IRS Elements: As the number of reflecting elements increased, the proposed method continued to closely track the ES performance, outperforming GS and RS by approximately 22% and 56%, respectively, for $100 \times 100$ element IRSs.
Sum Rate vs. Network Size: Even as the network area expanded (increasing path loss), the JBUA scheme maintained its superiority over greedy and random baselines, staying close to the ES performance.

Significance and Claims
The paper claims that the proposed dual-tier IRS framework provides a scalable blueprint for real-world 6G deployments, particularly for supporting massive IoT connectivity in challenging, blockage-prone environments. By targeting the FR3 "Golden Band" and utilizing a hybrid terrestrial-aerial architecture, the work addresses a critical gap in current literature which has largely focused on mmWave/THz bands and single-tier deployments.

The authors assert that their method offers a practical trade-off between performance and complexity, achieving near-optimal sum rates without the prohibitive computational cost of exhaustive search. The paper concludes that this research validates the effectiveness of multi-layer IRS architectures for reliable 6G deployments in the FR3 spectrum. Future work suggested by the authors includes expanding the architecture to include Low Earth Orbit (LEO) satellite segments and investigating hybrid beamforming under user mobility conditions (tracking, Doppler, and handover).