Multi-hop wireless networks can be optimized using directional antennas, as they allow for in-depth interference management and network topology optimization. This type of optimization involves ensuring high operational guarantees such as instantaneous connectivity, minimum SNRs and SINRs thresholds, and improved QoS. It simplifies tasks of future network layers and allows for more relaxed routing protocols and scheduling. However, attaining optimal performance via network configuration involves selecting an antenna orientation for each node to create a link with another node. This is challenging, especially when the process is carried out in real-time. To tackle this challenge, we present nodes2net, a Deep Neural Network (DNN) that is trained to imitate solved, ideal network instances. This approach uses nodes' positions as inputs and produces a set of links as output. By leveraging learning of patterns and theoretically driven properties, nodes2net can generate reliable network configuration solutions when dealing with new sets of node positions. It utilizes efficient neural network aggregation operators to facilitate and process information about the nodes, to finally produce the final solution as set of links. Our results demonstrate the competitive performance of this method.