⚡ electrical engineering

Is Repeater-Assisted Massive MIMO Compatible with Dynamic TDD?

This paper proposes a joint amplification and phase shift optimization framework for repeater-assisted massive MIMO networks in dynamic TDD systems, deriving spectral efficiency expressions and developing an algorithm that successfully balances desired signal amplification against the cross-link interference inherent in concurrent uplink and downlink transmissions.

Original authors: Martin Andersson, Anubhab Chowdhury, Erik G. Larsson

Published 2026-04-24

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Martin Andersson, Anubhab Chowdhury, Erik G. Larsson

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

The Big Picture: The "Super-Listener" Problem

Imagine a busy city with two neighborhoods. In one neighborhood, people are talking (sending data up), and in the other, people are listening (receiving data down).

In the old days, these neighborhoods took turns: one hour for talking, one hour for listening. This is called Static TDD. It's safe, but it's slow because half the time, everyone is just waiting.

Now, imagine we try to do both at the same time to save time. This is Dynamic TDD. It's much faster, but it creates a new problem: Noise.

The people talking in Neighborhood A accidentally shout over the people listening in Neighborhood B.
The people listening in Neighborhood B accidentally hear the shouting from Neighborhood A.

This is called Cross-Link Interference (CLI). It's like trying to hear your friend at a party while someone else is shouting right next to you.

Enter the Repeater: The "Megaphone"

To fix bad signal coverage, engineers use Repeaters. Think of a repeater as a smart megaphone or a super-ear.

It listens to a weak signal.
It makes it louder (amplifies it).
It sends it out again.

The Catch: A standard megaphone amplifies everything it hears. If you shout "Hello!" into a megaphone, it gets louder. But if a siren is also blaring nearby, the megaphone makes the siren louder too!

In our Dynamic TDD scenario, the repeater is in a tough spot. It wants to boost the signal for the person listening, but it also accidentally boosts the noise from the people talking in the other neighborhood. If the noise gets too loud, the listener can't hear anything at all.

The Paper's Solution: The "Tuning Knob"

The researchers asked: "Can we use these megaphones in a busy, noisy environment without making the noise worse?"

Their answer is Yes, but only if we tune the megaphone very carefully. They developed a new way to control the repeater with two specific settings:

Volume (Amplification): How loud should the signal be?
Phase Shift (The "Twist"): This is the tricky part. Imagine sound waves are like ocean waves. If a wave is at its peak, and you add another wave at its peak, they get huge. But if you add a wave at its "trough" (the bottom), they cancel each other out.
- The researchers figured out how to twist the signal slightly before amplifying it. This allows the repeater to boost the good signal while canceling out or minimizing the bad noise.

How They Did It (The "Dance")

They created a computer algorithm that acts like a dance partner. Here is how it works:

Step 1: The network sends a signal.
Step 2: The algorithm adjusts the repeater's "Volume" and "Twist" to make the signal as clear as possible.
Step 3: The network adjusts its own antennas to match the new signal.
Step 4: They repeat this dance over and over (a few times) until they find the perfect balance where the signal is loud and the noise is quiet.

What They Found (The Results)

They tested this in a simulation with two cells and one repeater. Here is what happened:

The Good News: When they tuned the repeater correctly, the speed of the internet (Spectral Efficiency) went up significantly.
- For the "Listening" side, speeds went up by about 20%.
- For the "Talking" side, speeds went up by a massive 70%.
The "Sweet Spot": The repeater works best when it is placed closer to the people using it, but not right on the border between the two neighborhoods. If it's too close to the border, it picks up too much noise from the other side.
The Trade-off: Sometimes, helping one side makes the other side slightly slower (like turning up the volume on your music might annoy your neighbor). However, the paper shows that the gains are huge, while the losses are tiny. It's a win-win overall.

The Bottom Line

This paper proves that Dynamic TDD (doing things at the same time) and Repeaters (boosting signals) can work together, but you can't just use a simple "on/off" switch.

You need a smart, adjustable repeater that knows how to twist the signal to cancel out the noise. If you do this, you get a much faster, more efficient 5G network without needing to build expensive new towers.

In short: They figured out how to tune a "smart megaphone" so it shouts the right message clearly, even while the neighborhood is noisy.

1. Problem Statement

The paper addresses the integration of active repeaters into Dynamic Time-Division Duplex (TDD) massive MIMO networks.

Context: While static TDD is common, Dynamic TDD allows access points (APs) to switch between Downlink (DL) and Uplink (UL) based on traffic demand, improving spectral efficiency. However, this creates Cross-Link Interference (CLI) where DL transmissions interfere with UL receivers (and vice versa).
The Conflict: Repeaters are active devices that amplify and retransmit signals to enhance coverage. In Dynamic TDD, a repeater inevitably amplifies not only the desired signals but also the CLI.
The Core Question: Is it possible to deploy repeaters in Dynamic TDD systems such that the gain in desired signal strength outweighs the amplification of interference? If so, how should the repeater gain (amplification and phase) be optimized to manage this trade-off?

2. System Model

The authors propose a two-cell system model operating on the same time-frequency resources:

Nodes: One DL AP serving $K$ users, one UL AP serving $J$ users, and a single dual-antenna repeater.
Channels: The model accounts for direct links, repeater-assisted links, and all interference sources:
- Multi-user interference.
- Inter-AP interference (DL AP $\to$ UL AP).
- Inter-user interference (UL user $\to$ DL user).
- Amplified noise and interference introduced by the repeater.
Repeater Operation: The repeater receives a composite signal, applies a complex-valued gain $\alpha_r = r_r e^{j\phi_r}$ (where $r_r$ is amplification and $\phi_r$ is phase shift), and retransmits it. The gain is constrained by a maximum output power limit.

3. Methodology

The paper develops a mathematical framework and an iterative optimization algorithm:

A. Spectral Efficiency (SE) Derivation

The authors derive closed-form expressions for the achievable Downlink and Uplink Spectral Efficiencies.

They treat the repeater as part of the channel, creating "composite channels" (direct + via repeater).
The Signal-to-Interference-plus-Noise Ratio (SINR) for both DL and UL is formulated as a rational function of the repeater's amplification gain ( $r_r$ ).
Crucially, the SINR expressions include terms for amplified CLI and noise, allowing for a precise analysis of the trade-off.

B. Optimization Algorithm

To maximize SE, the authors propose an iterative framework to jointly optimize the amplification gain ( $r_r$ ) and phase shift ( $\phi_r$ ):

Fixed Vectors: Given fixed precoding (DL) and combining (UL) vectors, the SINR is a ratio of quadratic polynomials in $r_r$ . The optimal $r_r$ is found by solving for the stationary points of this function or checking boundary constraints.
Phase Search: The optimal phase shift is found by searching a discrete set of values in $[0, 2\pi]$ .
Iterative Update: Since the optimal gain depends on the precoding/combining vectors (and vice versa), the algorithm alternates between:
- Optimizing the repeater gain ( $\alpha_r$ ) for fixed beamforming vectors.
- Updating the beamforming vectors (using Maximum Ratio Transmission/Combining based on the new composite channels) for the fixed gain.
- This repeats until convergence (typically 5–10 iterations).

4. Key Contributions

Theoretical Framework: First derivation of closed-form DL and UL SEs for repeater-assisted massive MIMO specifically under Dynamic TDD, explicitly modeling the amplification of cross-link interference.
Complex-Valued Gain Optimization: Unlike previous works that only optimized real-valued amplification, this paper optimizes the complex-valued gain (amplitude + phase). This allows the repeater to act as a phase shifter to mitigate interference, not just an amplifier.
Iterative Algorithm: A low-complexity algorithm that jointly optimizes repeater gain and beamforming vectors to maximize SE.
Compatibility Analysis: Provides specific conditions (regarding repeater location) under which repeaters improve performance in Dynamic TDD.

5. Numerical Results

Simulations were conducted with $M=16$ antennas and a single user per cell:

Performance Gains:
- The proposed optimization can boost DL SE by up to ~20% and UL SE by up to ~70% compared to a system without repeaters.
- Maximum gains occur when the repeater is located closer to the AP than the cell border (e.g., near the user but not at the edge).
Location Sensitivity: If the repeater moves too close to the cell border, it picks up more CLI than desired signal, causing diminishing returns or performance loss.
Phase Shift Importance: Jointly optimizing the phase shift yields significantly higher gains than optimizing amplification alone, particularly for users with lower baseline SE.
Trade-off: Optimizing the repeater for one direction (e.g., DL) may slightly reduce the SE in the opposite direction (UL), but the net system gain is overwhelmingly positive.

6. Significance and Conclusion

Feasibility: The paper confirms that repeater-assisted massive MIMO is compatible with Dynamic TDD, provided the repeater gain is carefully optimized.
Practical Impact: The results suggest that active repeaters (now included in 5G NR Release 18) can be effectively deployed in dynamic traffic scenarios to enhance coverage and capacity, provided they are placed strategically and their phase/amplitude are tuned to manage interference.
Future Work: The authors suggest extending the framework to multiple repeaters and jointly optimizing power control alongside repeater gain.

In summary, the paper demonstrates that while Dynamic TDD introduces complex interference challenges, a smart, phase-aware repeater optimization strategy can turn these repeaters into powerful tools for spectral efficiency enhancement rather than sources of degradation.