Abstract
Extra large-scale multiple-input multiple-output (XL-MIMO) is a key technology for future wireless communication systems. This paper considers the effects of visibility region (VR) at the base station (BS) in a non-stationary multi-user XL-MIMO scenario, where only partial antennas can receive users' signal. In time division duplexing (TDD) mode, we first estimate the VR at the BS by detecting the energy of the received signal during uplink training phase. The probabilities of two detection errors are derived and the uplink channel on the detected VR is estimated. In downlink data transmission, to avoid cumbersome Monte-Carlo trials, we derive a deterministic approximate expression for ergodic average energy efficiency (EE) with the regularized zero-forcing (RZF) precoding. In frequency division duplexing (FDD) mode, the VR is estimated in uplink training and then the channel information of detected VR is acquired from the feedback channel. Invoking approximate results, we propose an alternate optimization algorithm to design the detection threshold and the pilot length in both TDD and FDD modes. The numerical results reveal the impacts of VR estimation error on ergodic average EE and demonstrate the effectiveness of our proposed algorithm.
The Challenge of Spatial Non-Stationarity
As wireless communication moves toward 6G, Extra Large-scale MIMO (XL-MIMO) has emerged as a critical technology. By employing extremely large antenna arrays, base stations (BS) can serve a high density of users with unprecedented spectral efficiency. However, the physical scale of these arrays introduces a phenomenon known as spatial non-stationarity.
Unlike traditional MIMO, where it is assumed every antenna "sees" every user, XL-MIMO arrays are so large that obstacles or distance variations mean a user is only visible to a specific subset of the array. This subset is defined as the Visibility Region (VR). If a base station fails to accurately identify these regions, it wastes energy processing noise from "non-visible" antennas or suffers signal loss by ignoring "visible" ones.
Visibility Region Estimation and Error Modeling
The authors propose an energy-based detection method to estimate the VR during the uplink training phase. By measuring the energy of received pilot signals, the BS decides which antennas belong to a user's VR. This process is susceptible to two types of errors:
- Missed Detection: Failing to identify a visible antenna.
- False Detection: Mistakenly including a non-visible antenna.
A key insight of the paper is that missed detections are significantly more damaging to system performance than false detections. While false detections add a small amount of noise, missed detections directly weaken the intended signal power, drastically reducing the overall Energy Efficiency (EE).
Optimization for Maximum Energy Efficiency
To maximize the bits-per-Joule performance of the system, the paper addresses the trade-off involving pilot signal length ($\tau$) and the detection threshold ($\zeta_0$). A longer pilot improves VR estimation accuracy but leaves less time for actual data transmission within a coherence block.
To solve this, the authors derived "deterministic approximations"—complex mathematical formulas that predict average EE without the need for time-consuming Monte-Carlo simulations. Using these approximations, they developed an Alternate Optimization Algorithm to jointly find the optimal threshold and pilot length for both TDD and FDD modes.
Key Results and Insights
- Algorithm Efficiency: The proposed optimization algorithm is highly practical, typically converging to an optimal solution in fewer than 5 iterations.
- FDD Fragility: Frequency Division Duplexing (FDD) systems were found to be more sensitive to VR estimation errors than Time Division Duplexing (TDD) systems. In FDD, errors propagate into the feedback channel, wasting precious feedback bits on irrelevant antenna data.
- Heuristic Performance: Interestingly, a simple heuristic—setting the threshold to minimize the total sum of error probabilities—performs nearly as well as the complex optimal solution, offering a low-complexity path for hardware implementation.
- Model Validation: The derived deterministic formulas showed a perfect match with 10,000 Monte-Carlo trials, confirming the mathematical framework's robustness.
Conclusion
By accurately modeling and optimizing for spatial non-stationarity, this research provides a roadmap for more energy-efficient XL-MIMO systems. The ability to fine-tune antenna detection and training resources ensures that 6G networks can remain "green" while handling the massive data demands of the future. Future work will look to extend these models to more complex scenarios where user visibility regions only partially overlap.
Summary updated