Global stability/local optimality trade-offs in a class of decentralized integrated communication-navigation filters

Application of decentralized filters to improve the performance of relative navigation functions of an integrated communication-navigation system is discussed. The general form of the Supplemented Partitioning Approach (SPA) decentralized algorithm is derived. Trade-off criteria between global stability and local optimality of the decentralized system is introduced. It is found that the Riccati equation perturbation method seems to be a promising approach to investigate the global stability problem associated with decentralized filters.