Motivation:

Channel estimation is one of the fundamental problems of communication technologies. Often the channel estimate is based on a specific model which has the drawback that influences that are not incorporated in the model are not picked up by the estimation. To overcome this drawback we can think of a data driven approach which learns the underlying channel purely by given data.

Goal:

The aim of this thesis is to estimate the channel with a deep learning approach named Generative Adversarial Networks. To realize this, a fixed transmitter should be implemented and measurements should be made at the receiver side. Afterwards the Generative Adversarial Network is trained to learn a channel estimation using these measurements.

Requirements:

In order to process this thesis, knowledge of the lectures Wireless Communications Technologies, Communication Technologies and programming skills in Python or Matlab and GNU Radio are essential.