Method and device for extracting weak signals

Abstract

The invention provides a method and device for extracting weak signals. The method comprises the steps that sampling points from the kth to the (k+P-1)th in sampling signals are extracted, wherein the P is N or Ns, the N is a preset sampling signal length, the Ns is the preset length of the signal to be tested in the sampling signals, the k is an integer from 1 to (N-P+1), and the Ns is smaller than or equal to the N; a dictionary is established according to a preset frequency, a preset phase and a sampling point time; sparse representation vectors of corresponding signal values of the sampling points from the kth to the (k+P-1)th are solved according to the established dictionary; the filling signals or candidate signals to be tested of the signals to be tested in the sampling points from the kth to the (k+P-1)th are reconstructured according to a first dictionary and the solved sparse representation vectors; the starting positions of the signals to be tested are calculated according to the matching rate between the reconstructured filling signals or candidate signals to be tested of the signals to be tested and the signal values corresponding to the sampling points from the kth to the (k+P-1)th. According to the method and device for extracting the weak signals, the starting positions of the signals to be tested can be estimated accurately and fast.

Description

technical field [0001] The invention relates to the field of communication, in particular to a method and device for extracting weak signals. Background technique [0002] In the mobile communication network, radio frequency interference signals will cause troubles such as dropped calls, poor voice quality, and channel loss to mobile communications within the coverage area of ​​the base station; in the blind area of ​​mobile communications, the communication signal is very weak The signals overlap with each other, causing problems such as co-channel interference, which makes the mobile phone and BP machine unable to use normally. Target group feature recognition is one of the main applications in the field of big data research. Often due to strong noise interference and the existence of communication blind spots, the received weak signals cannot be accurately identified, resulting in missing or invalid information data. Therefore, in the background of strong noise, it is ne...

AI Technical Summary

Problems solved by technology

[0010] The above method mainly has the following defects: (1) It is extremely sensitive to the waveform complexity of the sampling signal, so it is mainly used in single-frequency periodic signal detection; (2) Its performance guarantee usually requires more data collection points, so the performance is limited. The improvement needs to be at the cost of sacrificing the amount of computation; (3) The prior knowledge of the system requires more, and almost all the information of the signal to be tested needs to be known in advance, including parameters such as frequency, phase, and amplitude, in order to obtain accurate results. limited applicability

Images