1: \begin{abstract}

2: 		Coordinated weighted sum-rate maximization in multicell MIMO networks with intra- and intercell interference and local channel state at the base stations is recognized as an important yet difficult problem.

3: 		A classical, locally optimal solution is obtained by the \gls{wmmse} algorithm which facilitates a distributed implementation in multicell networks.

4: 		However, it often suffers from slow convergence and therefore large communication overhead.

5: 		To obtain more practical solutions, the unrolling/unfolding of traditional iterative algorithms gained significant attention.

6: 		In this work, we demonstrate a complete unfolding of the \gls{wmmse} algorithm for transceiver design in multicell \acrshort{mu}-\acrshort{mimo} interference channels with local channel state information.

7: 		The resulting architecture termed GCN-WMMSE applies ideas from graph signal processing and is agnostic to different wireless network topologies, while exhibiting a low number of trainable parameters and high efficiency \wrt training data.

8: 		It significantly reduces the number of required iterations while achieving performance similar to the \gls{wmmse} algorithm, alleviating the overhead in a distributed deployment.

9: 		Additionally, we review previous architectures based on unrolling the \gls{wmmse} algorithm and compare them to GCN-WMMSE in their specific applicable domains.

10: 	\end{abstract}