Low-rank csi feedback method, storage medium and device for deep iterative neural network

Abstract

The invention discloses a low-rank CSI feedback method, storage medium and equipment for a deep iterative neural network. FISTA is expanded into a deep iterative neural network; the channel matrix is ​​decomposed based on the low rank of the channel to obtain the linearly independent part of the channel. and the rest; based on the linearly independent part of the channel and the rest of the channel, the linearly independent part of the reconstructed low-rank channel is reconstructed to obtain a better initial value of the rest; based on the sparsity and the rest of the low-rank channel The optimal initial value of , and the FISTA algorithm is used to expand the deep iterative neural network to reconstruct the linearly independent part and the rest of the low-rank channel respectively; use the CSI to compress the feedback information and the measurement matrix to train the network, and use the trained network to reconstruct the The linearly independent part and the rest of the low-rank channel reconstruct the complete low-rank CSI. The present invention reconstructs the linearly independent part and the remaining part of the low-rank channel matrix respectively through the low-rank property of the channel, so as to improve the CSI reconstruction accuracy.

Description

technical field [0001] The present invention belongs to the technical field, and in particular relates to a low-rank CSI feedback method, storage medium and device for a deep iterative neural network. Background technique [0002] Massive MIMO (Multi-input Multi-output) is one of the important technologies for 5G and future wireless communication systems. This technology deploys multiple transmit antennas and receive antennas at the signal receiving end and the transmitting end respectively. On the premise that the base station obtains accurate downlink CSI (Channel State Information), the spectrum efficiency of the communication system is higher and the system capacity is larger. , stronger robustness and so on. [0003] In a TDD (Time Division Duplex) system, according to the reciprocity of uplink and downlink channels, the base station can easily obtain downlink CSI. However, in an FDD (Frequency Division Duplex) system, since the uplink and downlink channels are in dif...

AI Technical Summary

Problems solved by technology

However, most of these methods are data-driven, not very interpretable, and are sensitive to data

Images