A Coalitional Game Theoretic Outlook on Distributed Adaptive Parameter Estimation

摘要: In this paper, the parameter estimation problem based on diffusion least-mean-squares strategies is analyzed from a coalitional game theoretical perspective. Specifically, while selfishly minimizing only their own mean-square costs, nodes in network form coalitions that benefit them. Due to its nature, modeled as nontransferable and two scenarios are studied, one where each node's payoff includes suitable accuracy criterion another which graph-based communication cost also considered. former scenario, we first analyze nonemptiness of core games corresponding traditional strategies, then, analysis extended recently proposed node-specific setting have overlapped but different interests. latter after formulating graph providing sufficient conditions for nonemptiness, propose distributed formation algorithm, merge-and-split approach, converges stable coalition structure.