A machine learning-based mimo link adaptive transmission method

Abstract

The invention discloses a MIMO link adaptive transmission method based on machine learning, uses unsupervised learning self-encoding algorithm for feature extraction and dimensionality reduction, introduces the idea of ​​deep learning, and can reduce the Feature Dimensions and Computational Complexity. The present invention utilizes the logistic regression algorithm to construct the mapping relationship between the channel state information and the transmission parameters, which is different from the previous fixed parametric model, and can be trained based on sample data, and can be more accurate when the quality of the data set is better and covers all states. The mapping relationship between channel state information and transmission parameters is well established, and compared with the traditional single equivalent signal-to-noise ratio, the channel state information can be more fully utilized. In addition, the present invention also performs CQI selection based on the channel matrix, and studies the MIMO link adaptive method based on machine learning through the channel matrix and noise variance, which is not restricted by receiver design and has universal applicability.

Description

technical field [0001] The invention relates to the technical field of wireless communication, in particular to a machine learning-based MIMO link adaptive transmission method. Background technique [0002] Rapid and significant fluctuations are an important feature of instantaneous channels, and dynamically adjusting transmission modes for changing channel conditions is the key to achieving stability and transmission efficiency in wireless systems. The adaptive transmission technology dynamically selects transmission parameters according to the current channel state information (CSI, Channel State Information) so as to achieve higher throughput while maintaining transmission stability. Multiple key transmission parameters need to be determined in current and future transmission systems, such as transmit power, modulation and coding scheme (MCS, Modulation and Coding Scheme), in multi-antenna transmission and multi-antenna reception (MIMO, Multiple-Input-Multiple-Output) sys...

AI Technical Summary

Problems solved by technology

However, due to the coupling of various transmission factors such as users, mode selection, precoding, etc., and with the application of massive MIMO, it is extremely challenging to achieve link adaptation in transmission systems

[0003] Traditional link adaptation is a one-dimensional problem corresponding to a single SNR (Signal to Noise Ratio) and MCS, but due to the high dimensionality of channel state information in multi-channel transmission, a single SNR cannot effectively describe MIMO or normal The channel state of OFDM (Orthogonal Frequency Division Multiplexing), that is, the average SNR cannot provide enough information to determine the ideal transmission mode

Traditional machine learning-based methods often use raw data as feature sets, without feature extraction and dimensionality reduction to remove redundant information, resulting in a significant increase in system complexity and dimensionality

Images