A cyclic prefix OFDM reception method based on model-driven deep learning

Abstract

The invention discloses a cyclic prefix OFDM receiving method based on model-driven deep learning, which includes: (1) converting the pilot signal y in the received signal y p Transform to obtain the frequency domain pilot signal, and initialize the least squares estimation with the local frequency domain pilot signal to obtain the least squares channel estimation result, and then input it to the first fully connected deep neural network for improvement, and perform the improved result The time-domain channel estimation signal is obtained by inverse Fourier transform (2) The received signal y is converted into a real-number domain received signal after eliminating inter-symbol interference (3) The real-number domain received signal is used as input, and the second depth is used according to the time-domain channel estimation signal The neural network performs iterative solution to obtain the final estimated modulation signal. (4) The transmission information bits are obtained after demodulating the modulation signal. The invention consumes less time and has high detection performance.

Description

technical field [0001] The present invention relates to communication technology, in particular to a model-driven deep learning-based OFDM receiving method without cyclic prefix. Background technique [0002] In recent years, deep learning, as a basic technology in artificial intelligence, has achieved great success in disciplines such as computer vision and natural language processing. Deep learning is a branch of the field of machine learning. It is a method of supervised learning. By minimizing the loss function between the predicted value and the real value of the deep neural network, a set of optimal neural network parameters is obtained to make the deep neural network The network is able to make accurate predictions. [0003] There have been some exploratory studies on the application of deep learning in the physical layer of wireless communication, including channel estimation, signal detection, encoder, decoder, channel feedback information reconstruction and end-to...

AI Technical Summary

Problems solved by technology

However, there are still very few examples of integrating the algorithmic knowledge in the field of wireless communication into the design of neural networks. Most of the functions of existing neural networks regard wireless communication systems or modules as a black box, and the training of neural networks depends entirely on a large number of Data-driven, the parameters of the neural network are relatively large, and the training speed is slow

In OFDM without cyclic prefix, inter-symbol interference and inter-carrier interference caused by multipath channels make channel estimation and channel detection face great challenges

Images