Complex channel equalizer design method based on complex value forward neural network

Abstract

The invention discloses a complex channel equalizer design method based on a complex value forward neural network, and relates to the field of artificial intelligence and communication. The complex channel equalizer design method based on the complex value forward neural network comprises the steps: employing a selectable search direction training method to train the complex value forward neural network, and enabling a target function value to be reduced as much as possible after each time of training, wherein the trained complex value forward neural network serves as a channel equalizer and is used in a digital communication system. The complex channel equalizer design method based on the complex value forward neural network has the beneficial effects that in the training process, a groupof direction factors are introduced to construct a plurality of search directions, the optimal training direction is selected from the search directions by adopting a breadth-first search strategy, and the optimal search direction is selected from the search directions. The high-efficiency training of the complex number forward neural network is completed, so that the design of the complex channel equalizer is realized, and excellent performance is obtained.

Description

technical field [0001] The invention relates to the fields of artificial intelligence and communication, in particular to a method for designing a complex channel equalizer based on a complex-valued forward neural network. Background technique [0002] With the rapid development of Internet technology, people's requirements for digital communication technology and digital communication quality are constantly improving, and efficient data transmission has become an important research hotspot. However, in the process of signal transmission, there are situations such as channel delay and multipath propagation, which inevitably lead to the degradation of communication quality. Therefore, when a signal passes through a wireless channel, the received signals will overlap each other, resulting in intersymbol interference. In addition, some other factors such as thermal noise, impulse noise and the nature of the channel itself will further lead to the degradation of communication q...

AI Technical Summary

Problems solved by technology

However, when the transmission channel is nonlinear or the noise interference is serious, the equalization performance of the linear equalizer will drop sharply

Images