Information Centrality Evaluation Method Based on Cascade Topological Relevance

Unmanned systems can be abstracted as dynamic and complex systems of multi-agent competition and cooperation. Its quantitative and qualitative characteristics are naturally similar to those in network science. Therefore, we can explore how to form a dynamic and efficient adjustment of the link relationship between nodes, based on studying of structural complexity, node complexity, and interactions between structure and nodes in network science. The aforementioned outputs can accordingly support the efficiency of information interaction and dissemination between nodes. To solve the problem of information cooperation in weak communication connection of unmanned systems, this paper proposed an information centrality evaluation method based on the degree of cascaded topology correlation (CTRICE, Cascade Topological Relevance Information Centrality Evaluation). The evaluation method and strategy of cascading topology association degree based on local neighborhood were formed through the evaluation of cascading information aggregation ability within the neighborhood and the evaluation of intimacy based on topology and interaction behavior. Consequently, the results of the assessment would provide support for information fusion and decision-making. This paper first proves the feasibility of this method in terms of information synergy consistency. Meanwhile, it compares and analyzes the convergence efficiency through simulation experiments between the proposed method with assessment methods of mean value and degree centrality. Compared with the traditional method, the proposed method has better robustness and robustness under the condition of low quality communication connections. The method presented in this paper provides an idea for the realization of information self-organization and collaboration based on topological relations in unmanned systems.