Browse Teach Curate

Model-Driven Deep Network for AoA Estimation via Spatial-Spectrum Image Reconstruction

A model-driven deep neural network that recasts angle-of-arrival estimation as reconstruction of a spatial-spectrum image, alternating a convolutional network with a sparse conjugate-gradient step to calibrate angular-dependent phase error in 5G antenna arrays.

Gort Text

Model-Driven Deep Neural Network for Enhanced AoA Estimation Using 5G gNB

High-accuracy positioning has become a fundamental enabler for intelligent connected devices. Nevertheless, the present wireless networks still rely on model-driven approaches to achieve positioning …