Deep Convolutional Neural Networks for Massive MIMO Fingerprint-Based Positioning
This work addresses positioning accuracy in wireless communication systems, specifically for massive MIMO applications, but it is incremental as it applies existing CNNs to a new domain with measured data.
The paper tackles fingerprint-based positioning using massive MIMO channels by applying convolutional neural networks (CNNs) to learn sparse channel structures, achieving fractional-wavelength positioning accuracies with moderately deep CNNs when sufficient training data is available.
This paper provides an initial investigation on the application of convolutional neural networks (CNNs) for fingerprint-based positioning using measured massive MIMO channels. When represented in appropriate domains, massive MIMO channels have a sparse structure which can be efficiently learned by CNNs for positioning purposes. We evaluate the positioning accuracy of state-of-the-art CNNs with channel fingerprints generated from a channel model with a rich clustered structure: the COST 2100 channel model. We find that moderately deep CNNs can achieve fractional-wavelength positioning accuracies, provided that an enough representative data set is available for training.