Reference signal extraction method and system based on minimum mean square error cancellation algorithm

Abstract

The invention discloses a reference signal extraction method and system based on a minimum mean square error cancellation algorithm. The method comprises the following steps: sampling a mixed signal;performing frequency hopping frequency point detection according to the sampled mixed signal to obtain a frequency hopping signal; intercepting a part of frequency hopping signals from the mixed signals; sampling the mixed signal for the second time according to part of the frequency hopping signal to obtain a mixed signal after sampling for the second time; sending part of the frequency hopping signals into a reference input end of a minimum mean square error cancellation algorithm model to serve as reference input signals; sending the mixed signal after the second sampling into a basic inputend of a minimum mean square error cancellation algorithm model to serve as a basic input signal; and performing filtering processing on the minimum mean square error cancellation algorithm model after the reference input signal and the basic input signal are input to obtain a reference signal. According to the invention, the efficiency and accuracy of reference signal extraction can be improved.

Description

technical field [0001] The invention relates to the field of reference signal extraction, in particular to a reference signal extraction method and system based on a minimum mean square error cancellation algorithm. Background technique [0002] One of the key technologies of the cancellation algorithm is to identify the reference signal for cancellation. The methods commonly used in academia are: neural network, system identification, blind signal modeling, etc. The problems with these methods are: the neural network method and system identification require high prior knowledge such as carrier frequency and phase, signal-to-noise ratio, etc. The large number of training samples makes it difficult to ensure real-time cancellation. The basis of blind signal modeling is to ensure that each signal is independent of each other. In actual situations, tracking interference signals are related to communication signals, and it is difficult to establish a signal model. Contents of ...

AI Technical Summary

Problems solved by technology

The methods commonly used in academia are: neural network, system identification, blind signal modeling, etc. The problems with these methods are: the neural network method and system identification require high prior knowledge such as carrier frequency and phase, signal-to-noise ratio, etc. The amount of training samples is large, and it is difficult to ensure real-time cancellation; the basis of blind signal modeling is to ensure that each signal is independent of each other. In actual situations, tracking interference signals and communication signals are related, and it is difficult to establish a signal model

Images