Abstract
We investigate Tx/Rx joint beamforming in millimeter-wave communications (MMWC). As the multipath components (MPCs) have different steering angles and independent fadings, beamforming aims at achieving array gain and diversity gain in this scenario. A sub-optimal beamforming scheme is proposed to find the antenna weight vectors (AWVs) at Tx/Rx via iterative eigenvalue decomposition (EVD), provided that full channel state information (CSI) is available at both the transmitter and receiver. To make this scheme practically feasible in MMWC, a corresponding training approach is suggested to avoid the channel estimation and iterative EVD computation. As in fast fading scenario, the training approach may be time-consuming due to frequent training; another beamforming scheme, which exploits the quasi-static steering angles in MMWC, is proposed to reduce the overhead and increase the system reliability by multipath grouping (MPG). The scheme first groups the MPCs and then concurrently beamforms toward multiple steering angles of the grouped MPCs, so that both array gain and diversity gain are achieved. Performance comparisons show that, compared with the corresponding state-of-the-art schemes, the iterative EVD scheme with the training approach achieves the same performance with a reduced overhead and complexity, whereas the MPG scheme achieves better performance with approximately equivalent complexity.
Introduction
Millimeter-wave (mmWave) communications, operating at frequencies such as 60 GHz, are pivotal for next-generation wireless systems due to their vast available bandwidth. However, these high frequencies suffer from significant propagation loss. To mitigate this, large antenna arrays are employed at both the transmitter (Tx) and receiver (Rx) to perform joint beamforming, focusing energy in specific directions to achieve high array gain.
A critical challenge in mmWave channels is their sparse but distinct multipath nature. Signals reflect off various surfaces, creating multiple multipath components (MPCs). While many systems focus on the single strongest path to maximize gain, this approach lacks robustness; if the primary path is obstructed, the link is lost. Achieving diversity gain—reliability through multiple paths—requires sophisticated beamforming that can target multiple MPCs simultaneously without the prohibitive overhead of full channel state information (CSI) estimation.
System Model
The system considers a point-to-point mmWave link where both Tx and Rx are equipped with antenna arrays. The channel is characterized by several MPCs, each with its own steering angle and fading coefficient.
The goal of joint beamforming is to determine the optimal Antenna Weight Vectors (AWVs), denoted as $\mathbf{w}_t$ and $\mathbf{w}_r$, to maximize the received signal power while potentially exploiting diversity across the paths $h_\ell$ and $g_\ell$.
Proposed Beamforming Schemes
1. Iterative EVD (IEVD) with Training
For scenarios with slow-fading channels, the authors propose an Iterative Eigenvalue Decomposition (IEVD) method. Instead of a computationally heavy singular value decomposition (SVD) of the entire channel matrix, IEVD uses an alternating optimization strategy. It fixes the receiver weights to optimize the transmitter weights, and then fixes the transmitter weights to optimize the receiver weights.
To enhance practical feasibility, a training approach based on the power method and channel reciprocity is introduced. This allows the Tx and Rx to converge to the optimal AWVs through a sequence of signal exchanges, bypassing the need for explicit channel estimation. This method significantly reduces overhead, requiring only $2(n_r + n_t)$ training sequences compared to the $n_r \times n_t$ required by traditional methods.
2. Multipath-Grouping (MPG)
In fast-fading environments, tracking the rapidly changing phases of individual MPCs is impractical. However, the spatial steering angles of these paths remain relatively stable (quasi-static). The Multipath-Grouping (MPG) scheme exploits this stability.
When the number of MPCs exceeds the number of antennas, traditional diversity-seeking beamformers often become mathematically unstable. MPG addresses this by grouping MPCs with similar angular characteristics into "equivalent" paths. By beamforming toward these grouped directions, the system achieves a balance between array gain and diversity gain, ensuring high reliability even in dense multipath environments.
Key Insights and Results
Simulation results using Monte Carlo methods demonstrate the effectiveness of the proposed schemes:
- Efficiency: The IEVD scheme typically reaches near-optimal performance in just 2 to 3 iterations, making it highly suitable for real-time applications.
- Robustness: The MPG scheme remains effective even when the number of paths ($L=20$) far exceeds the number of antennas (8), a scenario where previous diversity-seeking beamformers would mathematically fail.
- Trade-offs: The research highlights the inherent trade-off between diversity and array gain. As the number of paths increases, diversity improves reliability, but the average power gain per path decreases. The proposed schemes optimize this balance more effectively than existing solutions.
Conclusion
This paper presents two robust joint beamforming strategies for mmWave communications. The IEVD scheme provides a low-overhead solution for slow-fading channels via an iterative training approach, while the MPG scheme offers superior stability and performance in fast-fading, dense multipath scenarios. Future work may explore the integration of these techniques into hybrid beamforming architectures and validate performance on physical hardware testbeds.
Summary updated