Research On Signal Processing Methods For Multi-Base Cooperative Sensing Oriented To Perceptual Communication Integration

With the continuous integration of technologies such as the Internet of Things and artificial intelligence with traditional industries,new intelligent applications such as smart cities,intelligent transportation,and unmanned factories have emerged,and these applications urgently need new information infrastructure with functions such as perception,communication,and computing.come to support.As an important infrastructure supporting new applications such as smart cities and intelligent transportation,mobile communication systems are gradually combined with functions such as intelligent computing,and have evolved into infrastructure integrating perception,communication,and computing capabilities,which can support smart cities,intelligent transportation,etc.application.Base stations are an important type of equipment in wireless communication,and they are widely distributed.However,the signal power of base stations is limited and the field of view is limited.Therefore,in some cases,a single base station has poor perception of surrounding areas.In this case,the cooperation of multiple devices in the base station network can effectively make up for the lack of perception of a single base station.By fusing the signal data received by different base stations to extract the perception information of each base station on the environment/target,the fusion perception The result has higher accuracy than the single base station sensing result.The existing multi-node data fusion method mainly adopts the idea of"processing first,then fusion",that is,each sensing node first processes data according to the received signal,and uploads the data after obtaining relevant parameters about the target,and the fusion center The target parameters uploaded by different nodes are fused,and finally the perception results about the target are obtained,The disadvantage of this method is that there may be some correlation or complementary information between different received signals,and a single node greatly compresses the original information during information processing,thus losing the potential information correlation,so it cannot maximize the use of multiple signals.containing information for the best perceptual accuracy.In order to mine and utilize the correlation between different signals as much as possible to improve the accuracy of the sensing results,the multi-base station cooperative sensing method designed in this paper adopts the idea of "fusion first,then process".According to this relationship,the data extracted from multiple signals are adjusted and associated,so that the fusion operation of each signal data is carried out to achieve accurate perception of the target.This paper mainly designs fusion algorithms for multi-base station active sensing scenarios and passive sensing scenarios.In the multi-base station active sensing scenario,a method of using sign conjugate multiplication to eliminate the irrelevant phase information in the feature vector extracted from the signal is firstly designed,and the initial phase is 0 and only contains the target distance/velocity feature after reconstruction.symbol sequence;in addition,a multi-signal data fusion method is designed to highlight the real results by accumulating the distance or radial velocity difference between the grid points to be tested and the real results of the target relative to each base station,and improve the accuracy and resolution of target positioning and speed measurement results Rate.In the multi-base station active and passive sensing scenario,firstly,the accurate estimation of the phase of the central element of the vector is realized by improving the IDFT algorithm,and all elements of the vector are adjusted based on the phase of the central element to realize the phase synchronization of different vectors.Secondly,the distance feature vector of the active signal is used to construct a sequence related to the distance feature vector of the passive signal,which paves the way for subsequent correlation data fusion.Afterwards,the phase estimation of the first and last elements of the vector is realized by clipping and splicing the single distance feature vector,and multiple sequences are connected.Finally,on the basis of the phase synchronization of each vector,the joint function value of the grid points to be tested is calculated to determine the positioning result of the target.