eduzhai > Applied Sciences > Engineering >

High Dimensional Channel Estimation Using Deep Generative Networks

  • Save

... pages left unread,continue reading

Document pages: 30 pages

Abstract: This paper presents a novel compressed sensing (CS) approach to highdimensional wireless channel estimation by optimizing the input to a deepgenerative network. Channel estimation using generative networks relies on theassumption that the reconstructed channel lies in the range of a generativemodel. Channel reconstruction using generative priors outperforms conventionalCS techniques and requires fewer pilots. It also eliminates the need of apriori knowledge of the sparsifying basis, instead using the structure capturedby the deep generative model as a prior. Using this prior, we also performchannel estimation from one-bit quantized pilot measurements, and propose anovel optimization objective function that attempts to maximize the correlationbetween the received signal and the generator s channel estimate whileminimizing the rank of the channel estimate. Our approach significantlyoutperforms sparse signal recovery methods such as Orthogonal Matching Pursuit(OMP) and Approximate Message Passing (AMP) algorithms such as EM-GM-AMP fornarrowband mmWave channel reconstruction, and its execution time is notnoticeably affected by the increase in the number of received pilot symbols.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×