Distributed Filtering for Time-Varying Complex Networks under Energy Harvesting Constraints

This paper concerns the distributed filtering problem for a class of nonlinear complex networks with time-varying topologies and energy harvesting constraints. Each node is fitted with an energy harvester, which can acquire power from the outside. The energy harvesting level follows a first-order Markov model. The filter can access the measurement from the node when the energy amount is sufficient. The measurement transmission expectation can be calculated recursively. An upper bound of the estimation error covariance is presented, and this bound achieves its minimum by solving the optimal gain matrix. A simulation example confirms the effectiveness of the proposed estimation method.