Abstract
In time-division duplexing (TDD) systems, massive multiple-input multiple-output (MIMO) relies on the channel reciprocity to obtain the downlink (DL) channel state information (CSI) with the uplink (UL) CSI. In practice, the mismatches in the radio frequency (RF) analog circuits among different antennas at the base station (BS) break the end-to-end UL and DL channel reciprocity. Antenna calibration is necessary to avoid the severe performance degradation with massive MIMO. Many calibration schemes are available to compensate the RF gain mismatches and restore the channel reciprocity in TDD massive MIMO systems. In this paper, we focus on the internal self-calibration scheme where different BS antennas are interconnected via hardware transmission lines. First, we study the resulting calibration performance for an arbitrary interconnection strategy. Next, we obtain closed-form Cramer-Rao lower bound (CRLB) expressions for each interconnection strategy at the BS with only (M-1) transmission lines and M denotes the total number of BS antennas. Basing on the derived results, we further prove that the star interconnection strategy is optimal for internal self-calibration due to its lowest CRLB. In addition, we also put forward efficient recursive algorithms to derive the corresponding maximum-likelihood (ML) estimates of all the calibration coefficients. Numerical simulation results are also included to corroborate our theoretical analyses and results.
Introduction
Massive Multiple-Input Multiple-Output (MIMO) is a cornerstone of modern wireless communication, enabling high spectral efficiency through spatial multiplexing. In Time-Division Duplexing (TDD) systems, the principle of channel reciprocity is leveraged: the base station (BS) uses the uplink (UL) channel state information (CSI) to estimate the downlink (DL) channel. This eliminates the need for complex feedback from mobile stations (MS), as shown in the general system model.
However, while the physical propagation environment is reciprocal, the electronic hardware is not. The transmit (Tx) and receive (Rx) analog circuits at the BS exhibit different gains and phase shifts, which break the end-to-end reciprocity. Without precise calibration, the performance of beamforming and spectral efficiency degrades significantly.
The Challenge of RF Mismatch
Internal self-calibration is a robust method to address these hardware mismatches. In this approach, antennas at the BS are interconnected via physical transmission lines on the circuit board. These lines allow antennas to "sound" each other, measuring the combined effects of Tx/Rx mismatches and the calibration channel.
As illustrated in the figure above, the signal $y_{i,j}$ received at antenna $i$ from antenna $j$ depends on the Tx gain of antenna $j$ ($\alpha_j$), the Rx gain of antenna $i$ ($\beta_i$), and the reciprocal channel $h_{i,j}$. The goal is to estimate the calibration coefficients—the ratios of Tx to Rx gains—to restore reciprocity.
Evaluating Interconnection Strategies
A critical question arises: how should these $M$ antennas be wired together? Using the minimum number of connections ($M-1$), various topologies are possible, such as a "Daisy Chain" (linear) or a "Star" (all connected to one central reference).
The authors evaluate these strategies using the Cramer-Rao Lower Bound (CRLB), which defines the minimum possible variance for any unbiased estimator. They derived a closed-form expression for the CRLB of an arbitrary strategy: the variance for a specific antenna $m$ is proportional to $(d_m + 1)$, where $d_m$ is the number of "hops" or intermediate antennas between antenna $m$ and the reference antenna.
The Optimality of the Star Topology
Through mathematical proof using matrix transformations, the paper demonstrates that the Star interconnection strategy is optimal. In a star layout, every antenna is exactly one hop away from the reference, minimizing the total estimation error across the array. In contrast, a daisy chain suffers from error propagation, where antennas at the end of the chain exhibit much higher variance.
Key findings from numerical simulations include:
- MSE Performance: The proposed recursive Maximum Likelihood (ML) algorithms achieve the CRLB, confirming their efficiency.
- Spectral Efficiency: Systems calibrated using the star topology maintain near-ideal spectral efficiency even at lower Signal-to-Noise Ratios (SNR), significantly outperforming daisy-chain configurations.
Conclusion and Practical Implications
While the star topology is mathematically superior, the authors acknowledge the practical challenges of hardware implementation. Routing over 100 transmission lines to a single central point on a PCB can lead to physical congestion and signal integrity issues. Future work aims to explore "economical" wiring strategies that balance the theoretical optimality of the star layout with the physical constraints of circuit board design.
Summary updated