Deep Graph Unfolding for Beamforming in MU-MIMO Interference Networks
This work addresses beamforming optimization in wireless networks, offering an efficient solution for improving communication performance, though it is incremental as it builds on established WMMSE and unfolding techniques.
The paper tackles beamforming in multi-user MIMO interference networks by proposing unfolded WMMSE (UWMMSE), which combines the classical WMMSE method with graph neural networks, resulting in superior performance, generalizability, and robustness compared to existing methods.
We develop an efficient and near-optimal solution for beamforming in multi-user multiple-input-multiple-output single-hop wireless ad-hoc interference networks. Inspired by the weighted minimum mean squared error (WMMSE) method, a classical approach to solving this problem, and the principle of algorithm unfolding, we present unfolded WMMSE (UWMMSE) for MU-MIMO. This method learns a parameterized functional transformation of key WMMSE parameters using graph neural networks (GNNs), where the channel and interference components of a wireless network constitute the underlying graph. These GNNs are trained through gradient descent on a network utility metric using multiple instances of the beamforming problem. Comprehensive experimental analyses illustrate the superiority of UWMMSE over the classical WMMSE and state-of-the-art learning-based methods in terms of performance, generalizability, and robustness.