A privacy-preserving distributed credible evidence fusion algorithm for collective decision-making

By the fact that credible evidence fusion (CEF) does not satisfy data privacy and fusion consistency and hence cannot be extended to the distributed decision-making, a privacy-preserving credible evidence fusion (PCEF) algorithm with three-level consensus is proposed. In evidence difference measure (EDM) neighbor consensus, an evidence-free equivalent expression of EDM among neighboring agents is derived with the shared dot product protocol for pignistic probability and the identical judgment of two events with maximal subjective probabilities, so that evidence privacy is guaranteed due to such irreversible evidence transformation. In EDM network consensus, the neighbored EDMs reach uniformity via interaction between linear average consensus (LAC) while the EDMs between non-neighboring agents are inferred via low-rank matrix completion with rank adaptation to guarantee EDM consensus convergence and no solution for inferring raw evidence in numerical iteration style. In fusion network consensus, a privacy-preserving LAC with a self-cancelling random perturbation is proposed, where each agent adds its randomness to the sharing content and step-by-step cancels such randomness in consensus iterations. Furthermore, the sufficient condition for the convergence to the CCEF is explored, and it is proven that raw evidence cannot be uniquely reconstructed in such an iterative consensus. The comparison simulations show that PCEF is nearly approximate to CCEF both in credibility and fusion.