Fast Selection of Indoor Wireless Transmitter Locations with Generalizable Neural Network Propagation Models

The continuous emergence of new wireless communication systems increases the need for intelligent planning prior to their deployment. This planning includes determining the position of wireless access points, to meet quality of service objectives. How to effectively place a set of transmitters in an environment is a long-standing problem that has been widely studied over the years. This process has mainly relied on expensive measurement campaigns, low-fidelity empirical models, or high-fidelity but time-consuming simulations. Recent advances in scientific machine learning create new possibilities for overcoming this dichotomy between speed and accuracy. In this paper, we train a deep neural network (U-Net) to predict the received signal strength levels generated by multiple transmitters within a variety of different geometries. Then, we leverage the computational efficiency of the trained model to determine the position of transmitters in new geometries, both generated and real. This approach dramatically accelerates the process of selecting the position of access points, meeting multiple optimization objectives in an efficient manner. Also, for the first time, we experimentally demonstrate the viability of a U-Net based transmitter placement for an indoor Wi-Fi system.