Abstract: To address the issue of inconsistent estimates from multi-source sensors in distributed cooperative passive localization and tracking, this paper proposes a distributed consistency algorithm based on Fast Covariance Interaction （FCI）。 First, a multi-sensor temporal registration model is constructed using the Taylor formula and adaptive density clustering theory to resolve temporal asynchrony caused by multi-source asynchronous sampling. Subsequently, each sensor performs local estimation by integrating information contributions from neighboring sensors. Next, a theoretical relationship between the iteration number and the consensus error is derived based on a given steady-state consensus error, ensuring that the information group converges to consensus within a finite number of iterations. On this basis, each sensor only needs to perform one global fast covariance interaction to complete information fusion, significantly improving the efficiency and accuracy of fusion. Finally, the algorithm is applied to a multi-source distributed collaborative passive positioning system. Simulation results demonstrate that the proposed method can effectively achieve consensus fusion positioning within a finite number of iterations while simultaneously completing temporal alignment.

Key words: Multi-source Information Fusion, Distributed Consensus State Estimation, Cooperative Passive Localization, Fast Covariance Interaction, Temporal Alignment