Tensor Based Channel Parameter Estimation for Positioning Applications

This paper deduces a variational bayesian method for complex tensors based on mean field theory in MIMO-OFDM systems, which realizes automatic determination of tensor rank and applies CP decomposition of tensors to practical measurements through the tensor space structure. In this method, the channel harmonics are firstly estimated, and then used to implement the angle positioning method. To the end, the estimated harmonic errors and positioning errors are analyzed, confirming the achievability of this method.