Research On Structure Recognition Technology Of Wireless Communication Network

Wireless communication network plays an important role in modern warfare,and it is of great significance to the research on network reconnaissance and countermeasure technology.This dissertation focuses on this issue and carries out in-depth research.The main contents of this dissertation are as follows:(1)To solve the problem that it is difficult for non-cooperative parties to obtain the topology,a topology inference algorithm based on the time consecutive relationship of node signals and sparse vector reconstruction is proposed to realize the inference of the topology structure by non-cooperative parties.(2)To solve the problem of difficult reconnaissance of links without transmission information,two new link prediction indexes using local path information are proposed.A link prediction model based on density peak clustering is established to complete the results of topological inference.In addition,an adaptive moment estimation optimizer is proposed to improve network embedding model,which combines network hierarchical representation to realize the prediction of unknown links.(3)Aiming at the problem that there are many kinds of importance indexes and the selection is not objective enough,an important node recognition method based on multi-index fusion is proposed.Principal component analysis is used to select suitable indexes,entropy weight method and grey correlation method are used to determine the ranking of principal components,so as to avoid subjective judgment affecting the rationality of the selection of indexes and identify important nodes in the network more accurately.(4)Focused on the issue that the k-shell methods need global topology information and cannot be used in dynamic networks,an important node identification method for dynamic networks based on neighborhood first asynchronous h-index is proposed.The theory proves that method adapts to the change of network topology and converges to the ks value.Only the local information of neighbor nodes is used to identify the important nodes of dynamic network more quickly.The simulation experiments are carried out and the results of the experiments demonstrate that the algorithms above are effective.